{"owner":"withcoral","repo":"coral","hasSkills":true,"totalSkillsCount":83,"totalTokensCount":58653,"categories":["anthropic-skill","root-instruction","subagent-persona"],"hasMcp":false,"mcpConfig":null,"found":[".agents/skills/add-recipe/SKILL.md",".agents/skills/add-recipe/agents/openai.yaml",".agents/skills/ask-matt/PHASE-BOUNDARIES.md",".agents/skills/ask-matt/SKILL.md",".agents/skills/ask-matt/agents/openai.yaml",".agents/skills/code-review/SKILL.md",".agents/skills/code-review/agents/openai.yaml",".agents/skills/codebase-design/DEEPENING.md",".agents/skills/codebase-design/DESIGN-IT-TWICE.md",".agents/skills/codebase-design/SKILL.md",".agents/skills/codebase-design/agents/openai.yaml",".agents/skills/diagnosing-bugs/SKILL.md",".agents/skills/diagnosing-bugs/agents/openai.yaml",".agents/skills/domain-modeling/ADR-FORMAT.md",".agents/skills/domain-modeling/GLOSSARY-FORMAT.md",".agents/skills/domain-modeling/SKILL.md",".agents/skills/domain-modeling/agents/openai.yaml",".agents/skills/grill-me/SKILL.md",".agents/skills/grill-me/agents/openai.yaml",".agents/skills/grill-with-docs/SKILL.md",".agents/skills/grill-with-docs/agents/openai.yaml",".agents/skills/grilling/SKILL.md",".agents/skills/grilling/agents/openai.yaml",".agents/skills/handoff/SKILL.md",".agents/skills/handoff/agents/openai.yaml",".agents/skills/implement-spec/SKILL.md",".agents/skills/implement-spec/agents/openai.yaml",".agents/skills/implement/SKILL.md",".agents/skills/implement/agents/openai.yaml",".agents/skills/improve-codebase-architecture/HTML-REPORT.md",".agents/skills/improve-codebase-architecture/SKILL.md",".agents/skills/improve-codebase-architecture/agents/openai.yaml",".agents/skills/pr/CREDITS.md",".agents/skills/pr/SKILL.md",".agents/skills/pr/agents/openai.yaml",".agents/skills/prototype/LOGIC.md",".agents/skills/prototype/SKILL.md",".agents/skills/prototype/UI.md",".agents/skills/prototype/agents/openai.yaml",".agents/skills/research/SKILL.md",".agents/skills/research/agents/openai.yaml",".agents/skills/retro/SKILL.md",".agents/skills/retro/agents/openai.yaml",".agents/skills/show-me/SKILL.md",".agents/skills/show-me/agents/openai.yaml",".agents/skills/tdd/SKILL.md",".agents/skills/tdd/agents/openai.yaml",".agents/skills/tdd/mocking.md",".agents/skills/tdd/tests.md",".agents/skills/teach/GLOSSARY-FORMAT.md",".agents/skills/teach/LEARNING-RECORD-FORMAT.md",".agents/skills/teach/MISSION-FORMAT.md",".agents/skills/teach/RESOURCES-FORMAT.md",".agents/skills/teach/SKILL.md",".agents/skills/teach/agents/openai.yaml",".agents/skills/to-benchmark/EXAMPLE.md",".agents/skills/to-benchmark/KNOWN_ISSUES.md",".agents/skills/to-benchmark/SKILL.md",".agents/skills/to-benchmark/TEMPLATE.md",".agents/skills/to-questionnaire/SKILL.md",".agents/skills/to-questionnaire/agents/openai.yaml",".agents/skills/to-spec/SKILL.md",".agents/skills/to-spec/agents/openai.yaml",".agents/skills/to-tickets/SKILL.md",".agents/skills/to-tickets/agents/openai.yaml",".agents/skills/triage/AGENT-BRIEF.md",".agents/skills/triage/OUT-OF-SCOPE.md",".agents/skills/triage/SKILL.md",".agents/skills/triage/agents/openai.yaml",".agents/skills/wait-what/SKILL.md",".agents/skills/wait-what/agents/openai.yaml",".agents/skills/wayfinder/SKILL.md",".agents/skills/wayfinder/agents/openai.yaml",".agents/skills/wizard/SKILL.md",".agents/skills/wizard/agents/openai.yaml",".agents/skills/writing-for-agents/SKILL-MECHANICS.md",".agents/skills/writing-for-agents/SKILL.md",".agents/skills/writing-for-agents/agents/openai.yaml","AGENTS.md","apps/AGENTS.md","docs/agents/domain.md","docs/agents/issue-tracker.md","docs/agents/triage-labels.md"],"skills":{".agents/skills/add-recipe/SKILL.md":"---\nname: add-recipe\ndescription: Research an upstream service and write a Provider Recipe for it in apps/reef/recipes.\nargument-hint: \"Service name, e.g. notion\"\ndisable-model-invocation: true\n---\n\nWrite a **Provider Recipe** for the service the user names. A recipe states facts about one Upstream Service (its source, how to authenticate, its inputs and servers) so Reef can pre-fill Provider Setup. It lives at `apps/reef/recipes//recipe.yaml`, with an optional icon beside it.\n\nBefore you start, read:\n\n- `GLOSSARY.md`, the **Provider Recipe** entry.\n- ADRs 0002, 0003 and 0004 in `docs/adr/`.\n- `apps/reef/recipes/schema.json`: the recipe format. Its `description` fields are the field docs.\n- Example recipes: `linear` (MCP, OAuth discovery), `github` (OpenAPI, device code, public client) and `sentry` (region input, owner-configured client, servers).\n\n## 1. Check the id\n\nDerive the id from the service name: lowercase, `[a-z][a-z0-9_]*`. If `apps/reef/recipes//` exists, stop and ask the user whether to update that recipe or pick a new id.\n\n## 2. Find a source\n\nCoral reads two kinds of source:\n\n- An **OpenAPI 3.0 or 3.1** document, fetched once from a URL without following redirects. Swagger 2.0 fails.\n- An **MCP server over Streamable HTTP**. Standalone SSE and stdio servers fail.\n\nSearch the vendor's official docs for both. Record the URL of every doc page you rely on; the PR description cites them.\n\nDone when you hold at least one usable source URL, or have reported that none exists.\n\n## 3. Plan, then wait\n\nShow the user a short plan and wait for their approval. Cover each point:\n\n- **Recipes and ids.** One source: use the bare id. OpenAPI and MCP both: offer two recipes, `` for OpenAPI and `_mcp` for MCP. Give both the same `display_name` and set `service: ` in the MCP recipe, so the catalog shows both on one card and suggests one namespace (as `github_mcp` does). A self-hosted edition with its own host: offer `_self_hosted` as well.\n- **Source URLs**, and for a third-party document, the audit result and why you pinned it or not.\n- **Authentication options** and the `default` (see below).\n- **Scopes** for each OAuth option, or why it names none.\n- **Inputs and servers**, e.g. a region.\n- **Icon** source.\n- **Gaps**: anything the format cannot express.\n\nAuthentication rules:\n\n- Include every method the service supports that a recipe `method` can express, e.g. OAuth and an API key.\n- Default to OAuth when it needs no extra setup: `oauth_discovery` for MCP, or device code or authorization code with a public client. Otherwise default to the API key or personal access token.\n- Leave the OAuth client id out, so the owner configures one (as `sentry` does). Coral could register its own application later; list that as a follow-up.\n- Name scopes whenever the service reads them from the authorization request. This includes an `oauth_discovery` server that does not advertise them (as `github_mcp` does). Request every scope it offers, read and write, with a comment naming what each grants. Coral reads and writes through providers.\n- Name no scopes when the service supplies them another way: an `oauth_discovery` server that advertises them (as `hubspot` does), or a service that takes them from the app's settings (as `attio` does). Omit `scopes` for `oauth_discovery`. The other OAuth methods require the field, so write `scopes: []`. Add a comment saying why the recipe names none, and for app settings, list the scopes the owner must turn on.\n- If the only way to authenticate needs a method the format lacks, stop and report the gap and what a new method would need. Extending the format touches Reef, proto and Rust, and is out of scope here.\n\nModel regions and similar choices as an `inputs` enum inside one recipe.\n\n## 4. Write the recipe\n\nFor each approved recipe, write `apps/reef/recipes//recipe.yaml`. Start it with `# yaml-language-server: $schema=../schema.json`, then match the shape of the example recipes.\n\n- `description` follows the pattern \"Issues, events and releases from Sentry.\"\n- Write `label` and `description` text in the Reef copy style from `AGENTS.md`.\n- Comment anything a reader cannot tell from the YAML alone.\n\nFor the icon, use the vendor's official brand SVG from their brand or press page, or the Simple Icons one. If you find no clean SVG, or its terms are unclear, leave `icons` out and list it as a follow-up.\nName the logo under `icons.light`. Reef shows it on a white plate in the light theme and a dark plate in the dark theme. If it does not read on the dark plate, e.g. a black glyph, add the vendor's official or Simple Icons light variant under `icons.dark`, named `-dark.svg`. If there is no clean variant or its terms are unclear, leave `icons.dark` out and list it as a follow-up.\n\nDone when every approved recipe has its file and the plan's decisions all appear in it.\n\n## 5. Test\n\nRun, and fix until both pass:\n\n```sh\nnpm ci --prefix apps/reef # once, if apps/reef/node_modules is missing\n(cd apps/reef && npx vitest run --project server recipes.server)\nnpm run check --prefix apps/reef\n```\n\nDo not use `npm test -- recipes.server`. It also runs the Storybook tests, which match no file and exit 1.\n\nThe recipe test also enforces rules the schema cannot, such as no undeclared `{{input.x}}` and no unused credential fields.\n\nCheck each source directly:\n\n- **OpenAPI**: fetch the URL. Confirm a 200 with no redirect, and an `openapi` field of `3.0.x` or `3.1.x`.\n- **MCP**: POST an `initialize` request to the URL. Confirm a Streamable HTTP response, or a 401 that points at OAuth when the server needs sign-in.\n\nThen give the user steps for a live test, which needs their browser and account:\n\n1. Run `./dev run` (see `docs/development.md`).\n2. In Reef, open `/workspaces//providers/new?recipe=`.\n3. Sign in, publish the provider, and run a query against one of its tables.\n\n## 6. Summarise\n\nReport:\n\n- The files you wrote.\n- Test results, including the source checks.\n- Follow-ups: missing icon, OAuth client to register, the third-party audit result, format gaps.\n- A draft PR description that cites the vendor doc URLs.\n\nThen offer to open the PR. If the user accepts, create a branch named `/add--recipe`, using the user's usual branch prefix, and commit there.\n",".agents/skills/add-recipe/agents/openai.yaml":"interface:\n display_name: \"Add Recipe\"\n short_description: \"Write a Provider Recipe for a service\"\npolicy:\n allow_implicit_invocation: false\n",".agents/skills/ask-matt/PHASE-BOUNDARIES.md":"# Phase boundaries\n\nA **phase** is a chunk of work inside a session: the grilling, the implementation, the QA. The definition is fuzzy on purpose: a phase ends when you think *\"ok, we're done with that\"*.\n\nThe **phase boundary** is the gap between two phases, and it is the only place this decision belongs. Mid-phase there is no decision to make: continue, or split the work that's left into subagents. Compacting mid-phase makes the agent lose the thread.\n\n## The five options\n\n| Option | What it does |\n| ------------ | --------------------------------------------------------------- |\n| **Continue** | Stay in the session. No context switch at all. |\n| **`/clear`** | Empty the context window and start from nothing. |\n| **`/handoff`** | Write a portable markdown file and seed a session anywhere with it. |\n| **Subagent** | Send the task to its own context window and get a report back. |\n| **`/compact`** | Compress this context and seed a fresh session with the summary. |\n\n## The tree\n\nWork top to bottom at the boundary. The first **yes** wins.\n\n**1. Can you continue in this session?** Two things make the answer yes: the next phase needs this phase as a **primary source**, or you have enough [smart zone](https://www.aihero.dev/ai-coding-dictionary/smart-zone) left (~150k tokens) for the next phase to fit. Grilling → implementation is the standard yes: the implementation wants the reasoning verbatim, not a summary of it. Continue costs nothing and loses nothing, so rule it out before anything else.\n\n**2. Is the context irrelevant to what comes next?** Is everything in this session (the exploration, the decisions, the dead ends) disposable? If so, **`/clear`**. It is the cheapest move on the board: it takes no time and hands back the whole window. `/clear` also isn't terminal: the old session stays resumable.\n\nThe cost of getting this wrong is one-way. Clear a *relevant* context and you lose the **why** behind what you built, and no amount of reading the diff back gets it returned.\n\n**3. Do you need to hand off?** `/handoff` is narrow. You need it only when you are:\n\n- swapping to a **new harness** (Claude → Codex),\n- moving to a **new directory** or repo,\n- sending the work to a **colleague**,\n- or forking a side task you found **mid-phase** without derailing what you're doing.\n\nThat list is the whole clause. What `/handoff` buys is **portability**: a file that travels. If nothing is travelling, you don't need it.\n\n**4. Can the task be done AFK?** Is it scoped tightly enough to run with you away from the keyboard, no steering? Then send it to a **subagent** and leave this session untouched. Automated review is the standard case: the agent reads the diff and reports, and you aren't needed while it does.\n\n**5. Otherwise, `/compact`.** Relevant context, same harness, same directory, and you need to stay in the loop: this is where the tree lands, and it lands here often. Pass it an instruction (`/compact we're going to QA this area`) so the summary keeps what the next phase needs.\n\n`/compact` is the **default, not the first reach**. It sits at the bottom because the four questions above it are all cheaper or more precise. The failure mode when people start here is a fresh session that is confidently wrong about a decision the summary flattened.\n\n## Primary and secondary sources\n\nEvery move except **Continue** turns a **primary source** into a **secondary source**: the session as it happened, replaced by a summary of it. The trade is always the same shape:\n\n| Source | Information | Noise | Room to move |\n| --------------------------------- | ----------- | ----- | ------------ |\n| Primary (Continue) | Full | Lots | Little |\n| Secondary (`/compact`, `/handoff`) | Lossy | Less | Lots |\n\nThis is why question 1 comes first. You only pay the lossiness when staying costs more than it saves.\n\n## These are judgement calls\n\nThe questions are not objective: each has taste in it, and the same boundary can go two ways on two days. The value is in asking them **in order**, at the boundary rather than in the middle of the work.\n",".agents/skills/ask-matt/SKILL.md":"---\nname: ask-matt\ndescription: Ask which skill or flow fits your situation. A router over the skills in this repo.\ndisable-model-invocation: true\n---\n\n# Ask Matt\n\nYou don't remember every skill, so ask.\n\nBefore stating what a skill does or recommending a step be skipped, read that skill's SKILL.md: the summaries here are for orientation only.\n\nA **flow** is a path through the skills. Most paths run along one **main flow**, and two **on-ramps** merge onto it. Everything else is standalone, or a vocabulary layer that runs underneath.\n\n## The main flow: idea → ship\n\nThe route most work travels. You have an idea and want it built.\n\n1. **`/grill-with-docs`** sharpens the idea by interview. Start here whenever you are **working in a working directory**: it's stateful, retaining what it learns in `GLOSSARY.md` and ADRs. (No working directory? Use `/grill-me` instead, covered under Standalone. Both run the same `/grilling` primitive; `grill-with-docs` is the one that leaves a paper trail, which makes it the better of the two whenever a repo is there to leave it in.)\n2. **Branch: can you settle every question in conversation?** If a question needs a runnable answer (state, business logic, a UI you have to see), detour through a prototype, bridged by **`/handoff`** in both directions (a prototype lives in its own directory, which is exactly what `/handoff` is for; see Phase boundaries):\n - **`/handoff`** out, then open a fresh session against that file,\n - **`/prototype`** to answer the question with throwaway code,\n - **`/handoff`** back what you learned, and reference it from the original idea thread.\n3. **Branch: is this a multi-session build?**\n - **Yes** → **`/to-spec`** (turn the thread into a spec), then **`/to-tickets`** to split it into tracer-bullet tickets, each declaring its **blocking edges**. Then work the tickets one of two ways:\n - **`/implement`** per ticket, **`/clear`ing context between each one**. On a local tracker that's one file per ticket under `.scratch//issues/`, worked blockers-first by hand; on a real tracker the edges become native blocking links, so any ticket whose blockers are done can be grabbed. Each ticket is self-contained, so the last one's context is disposable.\n - **`/implement-spec`** for the whole spec in one run. It reads the tickets as a **task graph**, runs implementer subagents across the ready **frontier** in parallel, and lands everything on one **integration branch**. Reach for it when you'd rather orchestrate the build than drive each ticket yourself.\n - **No** → **`/implement`** right here, in the same context window.\n\n Either way, the code gets built by driving **`/tdd`** (one red-green slice at a time) and closes out with **`/code-review`**, a two-axis review (Standards + Spec) of the diff. `/implement` runs both per ticket; `/implement-spec`'s implementers each drive `/tdd`, and it runs one `/code-review` over the integration branch. Reach for **`/tdd`** on its own when you just want to build a concrete behaviour test-first without a full spec, and **`/code-review`** on its own whenever you want to review a branch or PR against a fixed point.\n\n When the work goes up as a pull request, **`/pr`** shapes the body: the smallest visual that shows the change, before/after evidence that it works, and a one-way or two-way door call. It's model-invoked, so the agent reaches for it whenever it writes a PR.\n\n4. **`/retro`** closes the loop. After a build, and especially one that went sideways, it looks back over the session and suggests changes to the agent's **environment**, not the code: navigation pointers, automated checks, the coding standards `/code-review` enforces, steering files, tooling. Mechanical mistakes become deterministic checks; judgement calls become coding standards. The next build then starts from a better environment.\n\n### Context hygiene\n\nKeep steps 1–3 in **one unbroken context window** (don't compact or clear until after `/to-tickets`) so the grilling, spec, and tickets all build on the same thinking. Each `/implement` then starts fresh, working from the ticket. Run `/retro` in the session it's looking back on, before you clear; after clearing, point it at that session's log instead.\n\nThe limit on this is the **[smart zone](https://www.aihero.dev/ai-coding-dictionary/smart-zone)**: the window (~150k tokens on state-of-the-art models) within which the model still reasons sharply. If a session approaches it before `/to-tickets`, don't push on degraded; `/compact` at the nearest phase boundary and carry on (see Phase boundaries).\n\n## On-ramps\n\nA starting situation that generates work, then merges onto the main flow.\n\n- **Bugs and requests piling up** → **`/triage`**. It moves issues through triage roles and produces agent-ready issues, which **`/implement`** later picks up.\n\n Triage is only for issues **you didn't create**: bug reports, incoming feature requests, anything that arrives raw. Tickets that `/to-tickets` produced are already agent-ready, so **don't triage them**.\n\n- **Something's broken** → **`/diagnosing-bugs`**. For the hard ones: the bug that resists a first glance, the intermittent flake, the regression that crept in between two known-good states. It refuses to theorise until it has a **tight feedback loop** (one command that already goes red on *this* bug), then fixes with a regression test. Once the fix is in, run **`/retro`** in the same session to ask what would have prevented the bug; where the real finding is that there's no good seam to lock it down, that's a job for **`/improve-codebase-architecture`**.\n\n- **A huge, foggy effort: a greenfield project or a huge feature build, too big for one session** → **`/wayfinder`**, the most cognitively demanding flow here. When the way from here to the destination isn't visible yet, it charts a **shared map** of **decision tickets** on the issue tracker and resolves them one at a time, producing **decisions, not deliverables**, until the fog is pushed back and the way is clear. Where **`/grill-with-docs`** sharpens an idea you can hold in one session, wayfinder is for the idea you can't, and it's slower and denser, so save it for exactly that, never a well-scoped feature.\n\n When the map clears, **it hands off, it doesn't build**: merge onto the main flow at **`/to-spec`**, which collapses the map's linked decisions into a buildable plan, then `/to-tickets` and `/implement` as usual. Looping the map straight into `/implement` skips that collapse and throws the linked detail away, so go straight to `/implement` only when the effort turned out genuinely small.\n\n## Codebase health\n\nNot feature work, just upkeep.\n\n- **`/improve-codebase-architecture`** runs whenever you have a spare moment to keep the codebase good for agents to operate in. It surfaces **deepening opportunities**; picking one _generates an idea_ you can take into the main flow at `/grill-with-docs`. It's the survey that finds the candidates; **`/codebase-design`** (below) is the bench you design the chosen one on.\n\n## Vocabulary underneath\n\nTwo model-invoked references that run *beneath* the other skills, each the single source of truth for its vocabulary. Reach for them directly when the **words**, not the process, are the problem; or let the skills above pull them in.\n\n- **`/domain-modeling`**: sharpen the project's *domain* language: challenge a fuzzy term, resolve an overloaded word (\"account\" doing three jobs), record a hard-to-reverse decision as an ADR. It's the active discipline `/grill-with-docs` drives to keep `GLOSSARY.md` a clean glossary.\n- **`/codebase-design`** is the deep-module vocabulary (module, interface, depth, seam, adapter, leverage, locality) for designing a module's *shape*: a lot of behaviour behind a small interface at a clean seam. `/tdd` and `/improve-codebase-architecture` both speak it.\n\n## Phase boundaries\n\nA **phase** is a chunk of work inside a session: the grilling, the implementation, the QA. At the **boundary** between two of them you have five options, and picking between them is the fuzziest decision in this whole map:\n\n- **Continue**: stay put. Costs nothing, loses nothing.\n- **`/clear`**: empty the window, when nothing here matters to what's next.\n- **`/handoff`** writes a portable markdown file. Narrow: only for a **new harness**, a **new directory**, a **colleague**, or forking a side task **mid-phase**. What it buys is portability.\n- **Subagent**: send a tightly-scoped task to its own window and get a report back.\n- **`/compact`** compresses this context and seeds a fresh session with it. The **default**, at the bottom of the tree rather than the first reach.\n\nRead [PHASE-BOUNDARIES.md](PHASE-BOUNDARIES.md) for the ordered tree: the five questions, the reasoning behind each branch, and why the primary-source cost makes **Continue** the one to rule out first. Make the decision **at** a boundary; mid-phase, continue or split the rest into subagents.\n\n## Standalone\n\nOff the main flow entirely.\n\n- **`/grill-me`**: the same relentless interview as `/grill-with-docs`, but **stateless**: it saves nothing locally and builds no `GLOSSARY.md`. Reach for it when you are **not working in a working directory** (sharpening a plan, a design, a piece of writing, anything with no repo under it). If you are in a working directory, use `/grill-with-docs` instead: it runs the same interview and leaves a paper trail, so it is strictly the better one.\n- **`/grilling`** is the interview primitive itself: rounds, the frontier, facts are the agent's job and decisions are yours. `/grill-me` and `/grill-with-docs` are the two named ways in, and `/triage`, `/wayfinder` and `/improve-codebase-architecture` all run it internally. Reach for it directly only when you want the interview with no wrapper around it.\n- **`/prototype`** is a small, throwaway program that answers one design question: does this state model feel right, or what should this UI look like. Throwaway is a constraint on how the code is written, not a promise to destroy it: the answer folds into the real code, and the prototype itself is kept as a **primary source** on a `prototype/` branch out of main, pointed at from the implementation issue. It's the detour in step 2 of the main flow, but reach for it any time a design question is hard to settle on paper.\n- **`/research`**: delegate reading legwork to a **background agent**: it investigates a question against **primary sources**, then leaves a cited Markdown file in the repo. Keep working while it reads. The file it produces is something to take *into* the main flow at `/grill-with-docs`, since research feeds the thinking rather than replacing it.\n- **`/to-questionnaire`** comes in when the thing blocking you isn't in your head or the codebase but in **someone else's**, and it writes them a questionnaire to fill in. It's the inverse of `/grill-me`: instead of interviewing you about the subject, it interviews you about the **send** (who it's going to, what you need back) and aims the questions at the gap. What comes back is material for `/grill-with-docs` or `/to-spec`.\n- **`/wizard`** is for the steps only a **human** can take: provisioning infrastructure, setting up credentials or CI secrets, clicking through an unfamiliar third-party dashboard, running a one-off migration or cutover. It generates an interactive bash script that opens each URL, captures each value, and writes it into `.env` and GitHub secrets, so the procedure stops being something you re-explain to an agent every time. Model-invoked, so the agent reaches for it the moment it hits a wall only you can pass. If the agent could just do it itself, it should; this is for where a human is genuinely in the loop.\n- **`/wait-what`** is the corrective for a message that didn't land. Use it mid-conversation, inside any other skill, and the agent re-pitches what it just said with the context you were missing, in plain English, using the `GLOSSARY.md` vocabulary. It works after the fact; `/grill-with-docs` is the upfront cure, because a shared language agreed early is what stops the jargon arriving at all.\n- **`/teach`**: learn a concept over multiple sessions, using the current directory as a stateful workspace.\n- **`/writing-for-agents`** is the reference for writing documents agents consume: skills, AGENTS.md, pointed-at docs.\n\n## Precondition\n\n**`/setup-matt-pocock-skills`**: run before your first engineering flow to configure the issue tracker, triage labels, and doc layout the other skills assume. Custom issue trackers also work.\n",".agents/skills/ask-matt/agents/openai.yaml":"interface:\n display_name: \"Ask Matt\"\n short_description: \"Find the right skill or workflow\"\npolicy:\n allow_implicit_invocation: false\n",".agents/skills/code-review/SKILL.md":"---\nname: code-review\ndescription: \"Review the changes since a fixed point (commit, branch, tag, or merge-base) along two axes: Standards (does the code follow this repo's documented coding standards?) and Spec (does the code match what the originating issue/spec asked for?). Runs both reviews in parallel sub-agents and reports them side by side. Use when the user wants to review a branch, a PR, work-in-progress changes, or asks to \\\"review since X\\\".\"\n---\n\nTwo-axis review of the diff between `HEAD` and a fixed point the user supplies:\n\n- **Standards**: does the code conform to this repo's documented coding standards?\n- **Spec**: does the code faithfully implement the originating issue / spec?\n\nBoth axes run as **parallel sub-agents** so they don't pollute each other's context, then this skill aggregates their findings.\n\nThe issue tracker should have been provided to you. If not, tell the user to run `/setup-matt-pocock-skills`.\n\n## Process\n\n### 1. Pin the fixed point\n\nWhatever the user said is the fixed point (a commit SHA, branch name, tag, `main`, `HEAD~5`, etc.). If they didn't specify one, ask for it.\n\nCapture the diff command once: `git diff ...HEAD` (three-dot, so the comparison is against the merge-base). Also note the list of commits via `git log ..HEAD --oneline`.\n\nBefore going further, confirm the fixed point resolves (`git rev-parse `) and the diff is non-empty. A bad ref or empty diff should fail here, not inside two parallel sub-agents.\n\n### 2. Identify the spec source\n\nLook for the originating spec, in this order:\n\n1. Issue references in the commit messages (`#123`, `Closes #45`, GitLab `!67`, etc.), fetched via the workflow in the tracker doc.\n2. A path the user passed as an argument.\n3. A spec file under `docs/`, `specs/`, or `.scratch/` matching the branch name or feature.\n4. If nothing is found, ask the user where the spec is. If they say there isn't one, the **Spec** sub-agent will skip and report \"no spec available\".\n\n### 3. Identify the standards sources\n\nSearch the repo for every file that documents how code should be written. When `CODING_STANDARDS.md` or `CONTRIBUTING.md` exists, it must be on the list.\n\nOn top of whatever the repo documents, the Standards axis always carries the **smell baseline** below: a fixed set of Fowler code smells (_Refactoring_, ch.3) that applies even when a repo documents nothing. Two rules bind it:\n\n- **The repo overrides.** A documented repo standard always wins; where it endorses something the baseline would flag, suppress the smell.\n- **Always a judgement call.** Each smell is a labelled heuristic (\"possible Feature Envy\"), never a hard violation. Like any standard here, skip anything tooling already enforces.\n\nEach smell reads *what it is* → *how to fix*; match it against the diff:\n\n- **Mysterious Name**: a function, variable, or type whose name doesn't reveal what it does or holds. → rename it; if no honest name comes, the design's murky.\n- **Duplicated Code**: the same logic shape appears in more than one hunk or file in the change. → extract the shared shape, call it from both.\n- **Feature Envy**: a method that reaches into another object's data more than its own. → move the method onto the data it envies.\n- **Data Clumps**: the same few fields or params keep travelling together (a type wanting to be born). → bundle them into one type, pass that.\n- **Primitive Obsession**: a primitive or string standing in for a domain concept that deserves its own type. → give the concept its own small type.\n- **Repeated Switches**: the same `switch`/`if`-cascade on the same type recurs across the change. → replace with polymorphism, or one map both sites share.\n- **Shotgun Surgery**: one logical change forces scattered edits across many files in the diff. → gather what changes together into one module.\n- **Divergent Change**: one file or module is edited for several unrelated reasons. → split so each module changes for one reason.\n- **Speculative Generality**: abstraction, parameters, or hooks added for needs the spec doesn't have. → delete it; inline back until a real need shows.\n- **Message Chains**: long `a.b().c().d()` navigation the caller shouldn't depend on. → hide the walk behind one method on the first object.\n- **Middle Man**: a class or function that mostly just delegates onward. → cut it, call the real target direct.\n- **Refused Bequest**: a subclass or implementer that ignores or overrides most of what it inherits. → drop the inheritance, use composition.\n\n### 4. Spawn both sub-agents in parallel\n\nIssue both sub-agent calls together, in the foreground, and aggregate the reports they return.\n\n**Standards sub-agent prompt** should include:\n\n- The full diff command and commit list.\n- The list of standards-source files you found in step 3, **plus the smell baseline from step 3** pasted in full (the sub-agent has no other access to it).\n- The brief: \"Report, per file/hunk where relevant, (a) every place the diff violates a documented standard: cite the standard (file + the rule); and (b) any baseline smell you spot: name it and quote the hunk. Distinguish hard violations from judgement calls: documented-standard breaches can be hard, but baseline smells are always judgement calls, and a documented repo standard overrides the baseline. Skip anything tooling enforces. Under 400 words.\"\n\n**Spec sub-agent prompt** should include:\n\n- The diff command and commit list.\n- The path or fetched contents of the spec.\n- The brief: \"Report: (a) requirements the spec asked for that are missing or partial; (b) behaviour in the diff that wasn't asked for (scope creep); (c) requirements that look implemented but where the implementation looks wrong. Quote the spec line for each finding. Under 400 words.\"\n\nIf the spec is missing, skip the Spec sub-agent and note this in the final report.\n\n### 5. Aggregate\n\nPresent the two reports under `## Standards` and `## Spec` headings, verbatim or lightly cleaned. Do **not** merge or rerank findings, because the two axes are deliberately separate (see _Why two axes_).\n\nEnd with a one-line summary: total findings per axis, and the worst issue _within each axis_ (if any). Don't pick a single winner across axes: that's the reranking the separation exists to prevent.\n\n## Why two axes\n\nA change can pass one axis and fail the other:\n\n- Code that follows every standard but implements the wrong thing → **Standards pass, Spec fail.**\n- Code that does exactly what the issue asked but breaks the project's conventions → **Spec pass, Standards fail.**\n\nReporting them separately stops one axis from masking the other.\n",".agents/skills/code-review/agents/openai.yaml":"interface:\n display_name: \"Code Review\"\n short_description: \"Review a diff on standards and spec\"\n",".agents/skills/codebase-design/DEEPENING.md":"# Deepening\n\nHow to deepen a cluster of shallow modules safely, given its dependencies. Assumes the vocabulary in [SKILL.md](SKILL.md): **module**, **interface**, **seam**, **adapter**.\n\n## Dependency categories\n\nWhen assessing a candidate for deepening, classify its dependencies. The category determines how the deepened module is tested across its seam.\n\n### 1. In-process\n\nPure computation, in-memory state, no I/O. Always deepenable: merge the modules and test through the new interface directly. No adapter needed.\n\n### 2. Local-substitutable\n\nDependencies that have local test stand-ins (PGLite for Postgres, in-memory filesystem). Deepenable if the stand-in exists. The deepened module is tested with the stand-in running in the test suite. The seam is internal; no port at the module's external interface.\n\n### 3. Remote but owned (Ports & Adapters)\n\nYour own services across a network boundary (microservices, internal APIs). Define a **port** (interface) at the seam. The deep module owns the logic; the transport is injected as an **adapter**. Tests use an in-memory adapter. Production uses an HTTP/gRPC/queue adapter.\n\nRecommendation shape: *\"Define a port at the seam, implement an HTTP adapter for production and an in-memory adapter for testing, so the logic sits in one deep module even though it's deployed across a network.\"*\n\n### 4. True external (Mock)\n\nThird-party services (Stripe, Twilio, etc.) you don't control. The deepened module takes the external dependency as an injected port; tests provide a mock adapter.\n\n## Seam discipline\n\n- **One adapter means a hypothetical seam. Two adapters means a real one.** Don't introduce a port unless at least two adapters are justified (typically production + test). A single-adapter seam is just indirection.\n- **Internal seams vs external seams.** A deep module can have internal seams (private to its implementation, used by its own tests) as well as the external seam at its interface. Don't expose internal seams through the interface just because tests use them.\n\n## Testing strategy: replace, don't layer\n\n- Old unit tests on shallow modules become waste once tests at the deepened module's interface exist; delete them.\n- Write new tests at the deepened module's interface. The **interface is the test surface**.\n- Tests assert on observable outcomes through the interface, not internal state.\n- Tests should survive internal refactors, since they describe behaviour, not implementation. If a test has to change when the implementation changes, it's testing past the interface.\n",".agents/skills/codebase-design/DESIGN-IT-TWICE.md":"# Design It Twice\n\nWhen the user wants to explore alternative interfaces for a chosen deepening candidate, use this parallel sub-agent pattern. Based on \"Design It Twice\" (Ousterhout): your first idea is unlikely to be the best.\n\nUses the vocabulary in [SKILL.md](SKILL.md): **module**, **interface**, **seam**, **adapter**, **leverage**.\n\n## Process\n\n### 1. Frame the problem space\n\nBefore spawning sub-agents, write a user-facing explanation of the problem space for the chosen candidate:\n\n- The constraints any new interface would need to satisfy\n- The dependencies it would rely on, and which category they fall into (see [DEEPENING.md](DEEPENING.md))\n- A rough illustrative code sketch to ground the constraints, not a proposal, just a way to make the constraints concrete\n\nShow this to the user, then immediately proceed to Step 2. The user reads and thinks while the sub-agents work in parallel.\n\n### 2. Spawn sub-agents\n\nSpawn 3+ sub-agents in parallel. Each must produce a **radically different** interface for the deepened module.\n\nPrompt each sub-agent with a separate technical brief (file paths, coupling details, dependency category from [DEEPENING.md](DEEPENING.md), what sits behind the seam). The brief is independent of the user-facing problem-space explanation in Step 1. Give each agent a different design constraint:\n\n- Agent 1: \"Minimize the interface: aim for 1–3 entry points max. Maximise leverage per entry point.\"\n- Agent 2: \"Maximise flexibility: support many use cases and extension.\"\n- Agent 3: \"Optimise for the most common caller: make the default case trivial.\"\n- Agent 4 (if applicable): \"Design around ports & adapters for cross-seam dependencies.\"\n\nInclude both [SKILL.md](SKILL.md) vocabulary and GLOSSARY.md vocabulary in the brief so each sub-agent names things consistently with the architecture language and the project's domain language.\n\nEach sub-agent outputs:\n\n1. Interface (types, methods, params, plus invariants, ordering, error modes)\n2. Usage example showing how callers use it\n3. What the implementation hides behind the seam\n4. Dependency strategy and adapters (see [DEEPENING.md](DEEPENING.md))\n5. Trade-offs: where leverage is high, where it's thin\n\n### 3. Present and compare\n\nPresent designs sequentially so the user can absorb each one, then compare them in prose. Contrast by **depth** (leverage at the interface), **locality** (where change concentrates), and **seam placement**.\n\nAfter comparing, give your own recommendation: which design you think is strongest and why. If elements from different designs would combine well, propose a hybrid. Be opinionated: the user wants a strong read, not a menu.\n",".agents/skills/codebase-design/SKILL.md":"---\nname: codebase-design\ndescription: Shared vocabulary for designing deep modules. Use when the user wants to design or improve a module's interface, find deepening opportunities, decide where a seam goes, make code more testable or AI-navigable, or when another skill needs the deep-module vocabulary.\n---\n\n# Codebase Design\n\nDesign **deep modules**: a lot of behaviour behind a small interface, placed at a clean seam, testable through that interface. Use this language and these principles wherever code is being designed or restructured. The aim is leverage for callers, locality for maintainers, and testability for everyone.\n\n## Glossary\n\nUse these terms exactly: don't substitute \"component,\" \"service,\" \"API,\" or \"boundary.\" Consistent language is the whole point.\n\n**Module**: anything with an interface and an implementation. Deliberately scale-agnostic: a function, class, package, or tier-spanning slice. _Avoid_: unit, component, service.\n\n**Interface**: everything a caller must know to use the module correctly: the type signature, but also invariants, ordering constraints, error modes, required configuration, and performance characteristics. _Avoid_: API, signature (too narrow, they refer only to the type-level surface).\n\n**Implementation**: what's inside a module, its body of code. Distinct from **Adapter**: a thing can be a small adapter with a large implementation (a Postgres repo) or a large adapter with a small implementation (an in-memory fake). Reach for \"adapter\" when the seam is the topic; \"implementation\" otherwise.\n\n**Depth**: leverage at the interface. The amount of behaviour a caller (or test) can exercise per unit of interface they have to learn. A module is **deep** when a large amount of behaviour sits behind a small interface, **shallow** when the interface is nearly as complex as the implementation.\n\n**Seam** _(Michael Feathers)_: a place where you can alter behaviour without editing in that place; the *location* at which a module's interface lives. Where to put the seam is its own design decision, distinct from what goes behind it. _Avoid_: boundary (overloaded with DDD's bounded context).\n\n**Adapter**: a concrete thing that satisfies an interface at a seam. Describes *role* (what slot it fills), not substance (what's inside).\n\n**Leverage**: what callers get from depth. More capability per unit of interface they learn. One implementation pays back across N call sites and M tests.\n\n**Locality**: what maintainers get from depth. Change, bugs, knowledge, and verification concentrate in one place rather than spreading across callers. Fix once, fixed everywhere.\n\n## Deep vs shallow\n\n**Deep module** = small interface + lots of implementation:\n\n```\n┌─────────────────────┐\n│ Small Interface │ ← Few methods, simple params\n├─────────────────────┤\n│ │\n│ Deep Implementation│ ← Complex logic hidden\n│ │\n└─────────────────────┘\n```\n\n**Shallow module** = large interface + little implementation (avoid):\n\n```\n┌─────────────────────────────────┐\n│ Large Interface │ ← Many methods, complex params\n├─────────────────────────────────┤\n│ Thin Implementation │ ← Just passes through\n└─────────────────────────────────┘\n```\n\nWhen designing an interface, ask:\n\n- Can I reduce the number of methods?\n- Can I simplify the parameters?\n- Can I hide more complexity inside?\n\n## Principles\n\n- **Depth is a property of the interface, not the implementation.** A deep module can be internally composed of small, mockable, swappable parts; they just aren't part of the interface. A module can have **internal seams** (private to its implementation, used by its own tests) as well as the **external seam** at its interface.\n- **The deletion test.** Imagine deleting the module. If complexity vanishes, it was a pass-through. If complexity reappears across N callers, it was earning its keep.\n- **The interface is the test surface.** Callers and tests cross the same seam. If you want to test *past* the interface, the module is probably the wrong shape.\n- **One adapter means a hypothetical seam. Two adapters means a real one.** Don't introduce a seam unless something actually varies across it.\n\n## Designing for testability\n\nGood interfaces make testing natural:\n\n1. **Accept dependencies, don't create them.**\n\n ```typescript\n // Testable\n function processOrder(order, paymentGateway) {}\n\n // Hard to test\n function processOrder(order) {\n const gateway = new StripeGateway();\n }\n ```\n\n2. **Return results, don't produce side effects.**\n\n ```typescript\n // Testable\n function calculateDiscount(cart): Discount {}\n\n // Hard to test\n function applyDiscount(cart): void {\n cart.total -= discount;\n }\n ```\n\n3. **Small surface area.** Fewer methods = fewer tests needed. Fewer params = simpler test setup.\n\n## Relationships\n\n- A **Module** has exactly one **Interface** (the surface it presents to callers and tests).\n- **Depth** is a property of a **Module**, measured against its **Interface**.\n- A **Seam** is where a **Module**'s **Interface** lives.\n- An **Adapter** sits at a **Seam** and satisfies the **Interface**.\n- **Depth** produces **Leverage** for callers and **Locality** for maintainers.\n\n## Rejected framings\n\n- **Depth as ratio of implementation-lines to interface-lines** (Ousterhout): rewards padding the implementation. We use depth-as-leverage instead.\n- **\"Interface\" as the TypeScript `interface` keyword or a class's public methods**: too narrow: interface here includes every fact a caller must know.\n- **\"Boundary\"**: overloaded with DDD's bounded context. Say **seam** or **interface**.\n\n## Going deeper\n\n- **Deepening a cluster given its dependencies**, see [DEEPENING.md](DEEPENING.md): dependency categories, seam discipline, and replace-don't-layer testing.\n- **Exploring alternative interfaces**, see [DESIGN-IT-TWICE.md](DESIGN-IT-TWICE.md): spin up parallel sub-agents to design the interface several radically different ways, then compare on depth, locality, and seam placement.\n",".agents/skills/codebase-design/agents/openai.yaml":"interface:\n display_name: \"Codebase Design\"\n short_description: \"Vocabulary for deep-module design\"\n",".agents/skills/diagnosing-bugs/SKILL.md":"---\nname: diagnosing-bugs\ndescription: Diagnosis loop for hard bugs and performance regressions. Use when the user says \"diagnose\"/\"debug this\", or reports something broken/throwing/failing/slow.\n---\n\n# Diagnosing Bugs\n\nA discipline for hard bugs. Skip phases only when explicitly justified.\n\nWhen exploring the codebase, read `GLOSSARY.md` (if it exists) to get a clear mental model of the relevant modules, and check ADRs in the area you're touching.\n\n## Redact\n\nThis skill has you show commands, outputs and captured artifacts. **Redact every secret first**: write `` in its place. Build loops against env vars, so the credential stays in the environment rather than in what you show. Captured artifacts carry auth headers: quote only the lines that carry the signal.\n\nIf the redacted output is not enough to diagnose the bug, say so and ask the user.\n\n## Phase 1: Build a feedback loop\n\n**This is the skill.** Everything else is mechanical. If you have a **tight** pass/fail signal for the bug (one that goes red on _this_ bug), you will find the cause; bisection, hypothesis-testing, and instrumentation all just consume it. If you don't have one, no amount of staring at code will save you.\n\nSpend disproportionate effort here. **Be aggressive. Be creative. Refuse to give up.**\n\n### Ways to construct one, in roughly this order\n\n1. **Failing test** at whatever seam reaches the bug: unit, integration, e2e.\n2. **Curl / HTTP script** against a running dev server.\n3. **CLI invocation** with a fixture input, diffing stdout against a known-good snapshot.\n4. **Headless browser script** (Playwright / Puppeteer) that drives the UI and asserts on DOM/console/network.\n5. **Replay a captured trace.** Save a real network request / payload / event log to disk; replay it through the code path in isolation.\n6. **Throwaway harness.** Spin up a minimal subset of the system (one service, mocked deps) that exercises the bug code path with a single function call.\n7. **Property / fuzz loop.** If the bug is \"sometimes wrong output\", run 1000 random inputs and look for the failure mode.\n8. **Bisection harness.** If the bug appeared between two known states (commit, dataset, version), automate \"boot at state X, check, repeat\" so you can `git bisect run` it.\n9. **Differential loop.** Run the same input through old-version vs new-version (or two configs) and diff outputs.\n10. **HITL bash script.** Last resort. If a human must click, drive _them_ with `scripts/hitl-loop.template.sh` so the loop is still structured. Captured output feeds back to you.\n\nBuild the right feedback loop, and the bug is 90% fixed.\n\n### Tighten the loop\n\nTreat the loop as a product. Once you have _a_ loop, **tighten** it:\n\n- Can I make it faster? (Cache setup, skip unrelated init, narrow the test scope.)\n- Can I make the signal sharper? (Assert on the specific symptom, not \"didn't crash\".)\n- Can I make it more deterministic? (Pin time, seed RNG, isolate filesystem, freeze network.)\n\nA 30-second flaky loop is barely better than no loop; a 2-second deterministic one is tight, a debugging superpower.\n\n### Non-deterministic bugs\n\nThe goal is not a clean repro but a **higher reproduction rate**. Loop the trigger 100×, parallelise, add stress, narrow timing windows, inject sleeps. A 50%-flake bug is debuggable; 1% is not, so keep raising the rate until it's debuggable.\n\n### When you genuinely cannot build a loop\n\nStop and say so explicitly. List what you tried. Ask the user for: (a) access to whatever environment reproduces it, (b) a redacted captured artifact (HAR file, log dump, core dump, screen recording with timestamps), or (c) permission to add temporary production instrumentation. Do **not** proceed to hypothesise without a loop.\n\n### Completion criterion: a tight loop that goes red\n\nPhase 1 is done when the loop is **tight** and **red-capable**: you can name **one command** (a script path, a test invocation, a curl) that you have **already run at least once** (show the invocation and its output, redacted), and that is:\n\n- [ ] **Red-capable**: it drives the actual bug code path and asserts the **user's exact symptom**, so it can go red on this bug and green once fixed. Not \"runs without erroring\"; it must be able to _catch this specific bug_.\n- [ ] **Deterministic**: same verdict every run (flaky bugs: a pinned, high reproduction rate, per above).\n- [ ] **Fast**: seconds, not minutes.\n- [ ] **Agent-runnable**: you can run it unattended; a human in the loop only via `scripts/hitl-loop.template.sh`.\n\nIf you catch yourself reading code to build a theory before this command exists, **stop: jumping straight to a hypothesis is the exact failure this skill prevents.** No red-capable command, no Phase 2.\n\n## Phase 2: Reproduce + minimise\n\nRun the loop. Watch it go red as the bug appears.\n\nConfirm:\n\n- [ ] The loop produces the failure mode the **user** described, not a different failure that happens to be nearby. Wrong bug = wrong fix.\n- [ ] The failure is reproducible across multiple runs (or, for non-deterministic bugs, reproducible at a high enough rate to debug against).\n- [ ] You have captured the exact symptom (error message, wrong output, slow timing) so later phases can verify the fix actually addresses it.\n\n### Minimise\n\nOnce it's red, shrink the repro to the **smallest scenario that still goes red**. Cut inputs, callers, config, data, and steps **one at a time**, re-running the loop after each cut, and keep only what's load-bearing for the failure.\n\nWhy bother: a minimal repro shrinks the hypothesis space in Phase 3 (fewer moving parts left to suspect) and becomes the clean regression test in Phase 5.\n\nDone when **every remaining element is load-bearing**: removing any one of them makes the loop go green.\n\nDo not proceed until you have reproduced **and** minimised.\n\n## Phase 3: Hypothesise\n\nGenerate **3–5 ranked hypotheses** before testing any of them. Single-hypothesis generation anchors on the first plausible idea.\n\nEach hypothesis must be **falsifiable**: state the prediction it makes.\n\n> Format: \"If is the cause, then will make the bug disappear / will make it worse.\"\n\nIf you cannot state the prediction, the hypothesis is a vibe: discard or sharpen it.\n\n**Show the ranked list to the user before testing.** They often have domain knowledge that re-ranks instantly (\"we just deployed a change to #3\"), or know hypotheses they've already ruled out. Cheap checkpoint, big time saver. Don't block on it; proceed with your ranking if the user is AFK.\n\n## Phase 4: Instrument\n\nEach probe must map to a specific prediction from Phase 3. **Change one variable at a time.**\n\nTool preference:\n\n1. **Debugger / REPL inspection** if the env supports it. One breakpoint beats ten logs.\n2. **Targeted logs** at the boundaries that distinguish hypotheses.\n3. Never \"log everything and grep\".\n\n**Tag every debug log** with a unique prefix, e.g. `[DEBUG-a4f2]`. Cleanup at the end becomes a single grep. Untagged logs survive; tagged logs die.\n\n**Perf branch.** For performance regressions, logs are usually wrong. Instead: establish a baseline measurement (timing harness, `performance.now()`, profiler, query plan), then bisect. Measure first, fix second.\n\n## Phase 5: Fix + regression test\n\nWrite the regression test **before the fix**, but only if there is a **correct seam** for it.\n\nA correct seam is one where the test exercises the **real bug pattern** as it occurs at the call site. If the only available seam is too shallow (single-caller test when the bug needs multiple callers, unit test that can't replicate the chain that triggered the bug), a regression test there gives false confidence.\n\n**If no correct seam exists, that itself is the finding.** Note it. The codebase architecture is preventing the bug from being locked down. Flag this for the next phase.\n\nIf a correct seam exists:\n\n1. Turn the minimised repro into a failing test at that seam.\n2. Watch it fail. If you forced the red by mutating code or a fixture, `diff` against a pristine copy to prove the mutation landed before you trust it.\n3. Apply the fix.\n4. Watch it pass.\n5. Re-run the Phase 1 feedback loop against the original (un-minimised) scenario.\n\n## Phase 6: Cleanup\n\nRequired before declaring done:\n\n- [ ] Original repro no longer reproduces (re-run the Phase 1 loop)\n- [ ] Regression test passes (or absence of seam is documented)\n- [ ] All `[DEBUG-...]` instrumentation removed (`grep` the prefix)\n- [ ] Throwaway prototypes deleted (or moved to a clearly-marked debug location)\n- [ ] The hypothesis that turned out correct is stated in the commit / PR message, so the next debugger learns\n",".agents/skills/diagnosing-bugs/agents/openai.yaml":"interface:\n display_name: \"Diagnosing Bugs\"\n short_description: \"Diagnose hard bugs and regressions\"\n",".agents/skills/domain-modeling/ADR-FORMAT.md":"# ADR Format\n\nADRs live in `docs/adr/` and use sequential numbering: `0001-slug.md`, `0002-slug.md`, etc.\n\nCreate the `docs/adr/` directory lazily: only when the first ADR is needed.\n\n## Template\n\n```md\n# {Short title of the decision}\n\n{1-3 sentences: what's the context, what did we decide, and why.}\n```\n\nThat's it. An ADR can be a single paragraph. The value is in recording *that* a decision was made and *why*, not in filling out sections.\n\n## Optional sections\n\nOnly include these when they add genuine value. Most ADRs won't need them.\n\n- **Status** frontmatter (`proposed | accepted | deprecated | superseded by ADR-NNNN`): useful when decisions are revisited\n- **Considered Options**: only when the rejected alternatives are worth remembering\n- **Consequences**: only when non-obvious downstream effects need to be called out\n\n## Numbering\n\nScan `docs/adr/` for the highest existing number and increment by one.\n\n## When to offer an ADR\n\nAll three of these must be true:\n\n1. **Hard to reverse**: the cost of changing your mind later is meaningful\n2. **Surprising without context**: a future reader will look at the code and wonder \"why on earth did they do it this way?\"\n3. **The result of a real trade-off**: there were genuine alternatives and you picked one for specific reasons\n\nIf a decision is easy to reverse, skip it: you'll just reverse it. If it's not surprising, nobody will wonder why. If there was no real alternative, there's nothing to record beyond \"we did the obvious thing.\"\n\n### What qualifies\n\n- **Architectural shape.** \"We're using a monorepo.\" \"The write model is event-sourced, the read model is projected into Postgres.\"\n- **Integration patterns between contexts.** \"Ordering and Billing communicate via domain events, not synchronous HTTP.\"\n- **Technology choices that carry lock-in.** Database, message bus, auth provider, deployment target. Not every library: just the ones that would take a quarter to swap out.\n- **Boundary and scope decisions.** \"Customer data is owned by the Customer context; other contexts reference it by ID only.\" The explicit no-s are as valuable as the yes-s.\n- **Deliberate deviations from the obvious path.** \"We're using manual SQL instead of an ORM because X.\" Anything where a reasonable reader would assume the opposite. These stop the next engineer from \"fixing\" something that was deliberate.\n- **Constraints not visible in the code.** \"We can't use AWS because of compliance requirements.\" \"Response times must be under 200ms because of the partner API contract.\"\n- **Rejected alternatives when the rejection is non-obvious.** If you considered GraphQL and picked REST for subtle reasons, record it; otherwise someone will suggest GraphQL again in six months.\n",".agents/skills/domain-modeling/GLOSSARY-FORMAT.md":"# GLOSSARY.md Format\n\n## Structure\n\n```md\n# {Context Name}\n\n{One or two sentence description of what this context is and why it exists.}\n\n## Language\n\n**Order**:\n{A one or two sentence description of the term}\n_Avoid_: Purchase, transaction\n\n**Invoice**:\nA request for payment sent to a customer after delivery.\n_Avoid_: Bill, payment request\n\n**Customer**:\nA person or organization that places orders.\n_Avoid_: Client, buyer, account\n```\n\n## Rules\n\n- **Be opinionated.** When multiple words exist for the same concept, pick the best one and list the others under `_Avoid_`.\n- **Keep definitions tight.** One or two sentences max. Define what it IS, not what it does.\n- **Only include terms specific to this project's context.** General programming concepts (timeouts, error types, utility patterns) don't belong even if the project uses them extensively. Before adding a term, ask: is this a concept unique to this context, or a general programming concept? Only the former belongs.\n- **Group terms under subheadings** when natural clusters emerge. If all terms belong to a single cohesive area, a flat list is fine.\n\n## Single vs multi-context repos\n\n**Single context (most repos):** One `GLOSSARY.md` at the repo root.\n\n**Multiple contexts:** A `GLOSSARY-MAP.md` at the repo root lists the contexts, where they live, and how they relate to each other:\n\n```md\n# Glossary Map\n\n## Contexts\n\n- [Ordering](./src/ordering/GLOSSARY.md): receives and tracks customer orders\n- [Billing](./src/billing/GLOSSARY.md): generates invoices and processes payments\n- [Fulfillment](./src/fulfillment/GLOSSARY.md): manages warehouse picking and shipping\n\n## Relationships\n\n- **Ordering → Fulfillment**: Ordering emits `OrderPlaced` events; Fulfillment consumes them to start picking\n- **Fulfillment → Billing**: Fulfillment emits `ShipmentDispatched` events; Billing consumes them to generate invoices\n- **Ordering ↔ Billing**: Shared types for `CustomerId` and `Money`\n```\n\nThe skill infers which structure applies:\n\n- If `GLOSSARY-MAP.md` exists, read it to find contexts\n- If only a root `GLOSSARY.md` exists, single context\n- If neither exists, create a root `GLOSSARY.md` lazily when the first term is resolved\n\nWhen multiple contexts exist, infer which one the current topic relates to. If unclear, ask.\n",".agents/skills/domain-modeling/SKILL.md":"---\nname: domain-modeling\ndescription: Build and sharpen a project's domain model. Use when discussing codebase terminology, writing or editing a GLOSSARY.md, or recording or editing an ADR.\n---\n\n# Domain Modeling\n\nActively build and sharpen the project's domain model as you design. This is the *active* discipline: challenging terms, inventing edge-case scenarios, and writing the glossary and decisions down the moment they crystallise. (Merely *reading* `GLOSSARY.md` for vocabulary is not this skill: that's a one-line habit any skill can do. This skill is for when you're changing the model, not just consuming it.)\n\n## File structure\n\nMost repos have a single context:\n\n```\n/\n├── GLOSSARY.md\n├── docs/\n│ └── adr/\n│ ├── 0001-event-sourced-orders.md\n│ └── 0002-postgres-for-write-model.md\n└── src/\n```\n\nIf a `GLOSSARY-MAP.md` exists at the root, the repo has multiple contexts. The map points to where each one lives:\n\n```\n/\n├── GLOSSARY-MAP.md\n├── docs/\n│ └── adr/ ← system-wide decisions\n├── src/\n│ ├── ordering/\n│ │ ├── GLOSSARY.md\n│ │ └── docs/adr/ ← context-specific decisions\n│ └── billing/\n│ ├── GLOSSARY.md\n│ └── docs/adr/\n```\n\nCreate files lazily: only when you have something to write. If no `GLOSSARY.md` exists, create one when the first term is resolved. If no `docs/adr/` exists, create it when the first ADR is needed.\n\n## During the session\n\n### Challenge against the glossary\n\nWhen the user uses a term that conflicts with the existing language in `GLOSSARY.md`, call it out immediately. \"Your glossary defines 'cancellation' as X, but you seem to mean Y. Which is it?\"\n\n### Sharpen fuzzy language\n\nWhen the user uses vague or overloaded terms, propose a precise canonical term. \"You're saying 'account': do you mean the Customer or the User? Those are different things.\"\n\n### Discuss concrete scenarios\n\nWhen domain relationships are being discussed, stress-test them with specific scenarios. Invent scenarios that probe edge cases and force the user to be precise about the boundaries between concepts.\n\n### Cross-reference with code\n\nWhen the user states how something works, check whether the code agrees. If you find a contradiction, surface it: \"Your code cancels entire Orders, but you just said partial cancellation is possible. Which is right?\"\n\n### Update GLOSSARY.md inline\n\nWhen a term is resolved, update `GLOSSARY.md` right there. Don't batch these up: capture them as they happen. Use the format in [GLOSSARY-FORMAT.md](./GLOSSARY-FORMAT.md).\n\n`GLOSSARY.md` should be totally devoid of implementation details. Do not treat `GLOSSARY.md` as a spec, a scratch pad, or a repository for implementation decisions. It is a glossary and nothing else.\n\n### Offer ADRs sparingly\n\nOnly offer to create an ADR when all three are true:\n\n1. **Hard to reverse**: the cost of changing your mind later is meaningful\n2. **Surprising without context**: a future reader will wonder \"why did they do it this way?\"\n3. **The result of a real trade-off**: there were genuine alternatives and you picked one for specific reasons\n\nIf any of the three is missing, skip the ADR. Use the format in [ADR-FORMAT.md](./ADR-FORMAT.md).\n",".agents/skills/domain-modeling/agents/openai.yaml":"interface:\n display_name: \"Domain Modeling\"\n short_description: \"Build and sharpen a domain model\"\n",".agents/skills/grill-me/SKILL.md":"---\nname: grill-me\ndescription: A relentless interview to sharpen a plan or design.\ndisable-model-invocation: true\n---\n\nCall the Skill tool with \"grilling\".\n",".agents/skills/grill-me/agents/openai.yaml":"interface:\n display_name: \"Grill Me\"\n short_description: \"Sharpen a plan through interview\"\npolicy:\n allow_implicit_invocation: false\n",".agents/skills/grill-with-docs/SKILL.md":"---\nname: grill-with-docs\ndescription: A relentless interview to sharpen a plan or design, which also creates docs (ADR's and glossary) as we go.\ndisable-model-invocation: true\n---\n\nCall the Skill tool twice, for \"grilling\" and \"domain-modeling\".\n",".agents/skills/grill-with-docs/agents/openai.yaml":"interface:\n display_name: \"Grill with Docs\"\n short_description: \"Grill a design and write its docs\"\npolicy:\n allow_implicit_invocation: false\n",".agents/skills/grilling/SKILL.md":"---\nname: grilling\ndescription: Grill the user relentlessly about a plan, decision, or idea. Use when the user wants to stress-test their thinking, or uses any 'grill' trigger phrases.\n---\n\nInterview the user relentlessly until you reach a shared understanding. Map this as a **design tree**: every decision branches into the decisions that hang off it.\n\nWork the tree in **rounds**. The **frontier** is every decision whose prerequisites are already settled: the questions you can ask _now_ without guessing at answers you haven't heard yet. Ask the whole frontier in one round: number each question and give your recommended answer. Then wait for the user's answers before the next round.\n\nFormat a round like so:\n\n```\n❓ **Q1** - ****: \n\n➡️ \n\n---\n\n❓ **Q2** - ****: \n\n➡️ \n```\n\nWord each question so \"yes\" accepts your recommended answer.\n\nEach round the user answers reshapes the tree: settled decisions push the frontier outward and unblock questions that depended on them. Recompute the frontier and ask the next round. A question whose answer depends on another question still open in this round belongs to a _later_ round, not this one.\n\nFinding _facts_ is your job, never the user's. When a frontier question needs a fact from the environment (filesystem, tools, etc.), dispatch a sub-agent to find it; don't ask the user for anything you could look up yourself. Don't block on it: a running exploration is an unsettled prerequisite, so only the questions downstream of it wait for the sub-agent to report; ask the rest of the frontier now. The _decisions_ are the user's: put each to them and wait.\n\nThe session is done when the frontier is empty: every branch of the design tree visited, nothing left silently assumed. Do not act on it until the user confirms you have reached a shared understanding.\n",".agents/skills/grilling/agents/openai.yaml":"interface:\n display_name: \"Grilling\"\n short_description: \"Stress-test thinking a round of questions at a time\"\n",".agents/skills/handoff/SKILL.md":"---\nname: handoff\ndescription: Compact the current conversation into a handoff document for another agent to pick up.\nargument-hint: \"What will the next session be used for?\"\ndisable-model-invocation: true\n---\n\nWrite a handoff document summarising the current conversation so a fresh agent can continue the work. Save to the temporary directory of the user's OS (`$TMPDIR`, else `/tmp`; `%TEMP%` on Windows) - not the current workspace.\n\nInclude a \"suggested skills\" section in the document, naming which skills the next agent should call the Skill tool for.\n\nDo not duplicate content already captured in other artifacts (specs, plans, ADRs, issues, commits, diffs). Reference them by path or URL instead.\n\nRedact any sensitive information, such as API keys, passwords, or personally identifiable information.\n\nIf the user passed arguments, treat them as a description of what the next session will focus on and tailor the doc accordingly.\n",".agents/skills/handoff/agents/openai.yaml":"interface:\n display_name: \"Handoff\"\n short_description: \"Compact a conversation into a handoff\"\npolicy:\n allow_implicit_invocation: false\n",".agents/skills/implement-spec/SKILL.md":"---\nname: implement-spec\ndescription: \"Implement the result of /to-spec and /to-tickets in code.\"\ndisable-model-invocation: true\n---\n\nYou have been provided a spec. This spec should have tickets associated with it, describing how to implement the spec.\n\nThe issue tracker should have been provided to you. If not, tell the user to run `/setup-matt-pocock-skills`.\n\nThe goal is the entire spec implemented on a single **integration branch**, with every ticket resolved the way the issue tracker closes work.\n\nThe tickets are not a list of steps. They are a **task graph** with blocking relationships between them. This means there is always a **frontier** of tickets which are ready to be grabbed.\n\nCommunication to and from subagents should be sparse. Communicate primarily through **context pointers**: to the spec, tickets, research notes, and previous commits. Don't duplicate information already available via pointers.\n\n**Implementer subagents** should be run in the background where possible for maximum concurrency.\n\n## Steps\n\n1. Read the spec and tickets to understand the task graph.\n\n2. (optional) Use an **exploration subagent** to conduct any exploration required by the tickets - relevant codebase files or external documentation. Ensure the exploration subagent can save files - it should save its markdown notes in a directory outside the repo, accessible by all future subagents. This lets **implementer subagents** focus on implementation rather than exploration.\n\n3. Create the integration branch. If the issue tracker closes work through PRs, or the user asks for one, open a draft PR after the first merge in step 5 (a branch with no commits ahead of main can't open one), marked as closing the spec and tickets.\n\n4. Use **implementer subagents** to implement each ticket, each in its own worktree on its own branch. Each implementer subagent:\n - confirms its worktree is based on the integration branch before starting, and resets onto it if not;\n - calls the Skill tool with `tdd` to build the ticket;\n - merges the integration branch tip into its own branch before reporting done\n\n5. Once an **implementer subagent** completes, merge its work to the integration branch with a **merger subagent**.\n\n6. If this changes the **frontier** of available tickets, kick off more **implementer subagents** to work on the new tickets. This allows for maximum concurrency.\n\n7. Once all tickets are complete, call the Skill tool with `code-review` on the integration branch. Fix all issues raised by the code review in a single **implementer subagent**.\n\n8. If a draft PR exists, mark it ready for review. Otherwise, resolve each ticket the way the issue tracker closes work, and report the integration branch.\n\n9. Clean up all **implementer subagent** worktrees.\n",".agents/skills/implement-spec/agents/openai.yaml":"interface:\n display_name: \"Implement Spec\"\n short_description: \"Implement the result of /to-spec and /to-tickets in code.\"\npolicy:\n allow_implicit_invocation: false\n",".agents/skills/implement/SKILL.md":"---\nname: implement\ndescription: \"Implement a piece of work based on a spec or set of tickets.\"\ndisable-model-invocation: true\n---\n\nImplement the work described by the user in the spec or tickets.\n\nIf the user passes a ticket reference, fetch it from the issue tracker and state its title before starting. If the reference is ambiguous, ask.\n\nCall the Skill tool with \"tdd\" where possible, at pre-agreed seams.\n\nRun typechecking regularly, single test files regularly, and the full test suite once at the end.\n\nOnce done, call the Skill tool with \"code-review\" to review the work.\n\nCommit your work to the current branch.\n",".agents/skills/implement/agents/openai.yaml":"interface:\n display_name: \"Implement\"\n short_description: \"Build work from a spec or tickets\"\npolicy:\n allow_implicit_invocation: false\n",".agents/skills/improve-codebase-architecture/HTML-REPORT.md":"# HTML Report Format\n\nThe architectural review is rendered as a single self-contained HTML file in the OS temp directory. Tailwind and Mermaid both come from CDNs. Mermaid handles graph-shaped diagrams reliably; hand-built divs and inline SVG handle the more editorial visuals (mass diagrams, cross-sections). Mix the two: don't lean on Mermaid for everything, it'll start to look generic.\n\n## Scaffold\n\n```html\n\n\n \n \n Architecture review for {{repo name}}\n \n \n \n \n \n
\n
...
\n
...
\n
...
\n
\n \n\n```\n\n## Header\n\nRepo name, date, and a compact legend: solid box = module, dashed line = seam, red arrow = leakage, thick dark box = deep module. No introduction paragraph. Straight into the candidates.\n\n## Candidate card\n\nThe diagrams carry the weight. Prose is sparse, plain, and uses the glossary terms (from the `/codebase-design` skill) without ceremony.\n\nEach candidate is one `
`:\n\n- **Title**: short, names the deepening (e.g. \"Collapse the Order intake pipeline\").\n- **Badge row**: recommendation strength (`Strong` = emerald, `Worth exploring` = amber, `Speculative` = slate), plus a tag for the dependency category (`in-process`, `local-substitutable`, `ports & adapters`, `mock`).\n- **Files**: monospaced list, `font-mono text-sm`.\n- **Before / After diagram**: the centrepiece. Two columns, side by side. See patterns below.\n- **Problem**: one sentence. What hurts.\n- **Solution**: one sentence. What changes.\n- **Wins**: bullets, ≤6 words each. e.g. \"Tests hit one interface\", \"Pricing logic stops leaking\", \"Delete 4 shallow wrappers\".\n- **ADR callout** (if applicable): one line in an amber-tinted box.\n\nNo paragraphs of explanation. If the diagram needs a paragraph to be understood, redraw the diagram.\n\n## Diagram patterns\n\nPick the pattern that fits the candidate. Mix them. Don't make every diagram look the same. Variety is part of the point.\n\n### Mermaid graph (the workhorse for dependencies / call flow)\n\nUse a Mermaid `flowchart` or `graph` when the point is \"X calls Y calls Z, and look at the mess.\" Wrap it in a Tailwind-styled card so it doesn't feel parachuted in. Style with classDef to colour leakage edges red and the deep module dark. Sequence diagrams work well for \"before: 6 round-trips; after: 1.\"\n\n```html\n
\n 
\n    flowchart LR\n      A[OrderHandler] --> B[OrderValidator]\n      B --> C[OrderRepo]\n      C -.leak.-> D[PricingClient]\n      classDef leak stroke:#dc2626,stroke-width:2px;\n      class C,D leak\n
\n
\n```\n\n### Hand-built boxes-and-arrows (when Mermaid's layout fights you)\n\nModules as `
`s with borders and labels. Arrows as inline SVG `` or `` elements positioned absolutely over a relative container. Reach for this when you want the \"after\" diagram to feel like one thick-bordered deep module with greyed-out internals, since Mermaid won't render that with the right weight.\n\n### Cross-section (good for layered shallowness)\n\nStack horizontal bands (`h-12 border-l-4`) to show layers a call passes through. Before: 6 thin layers each doing nothing. After: 1 thick band labelled with the consolidated responsibility.\n\n### Mass diagram (good for \"interface as wide as implementation\")\n\nTwo rectangles per module: one for interface surface area, one for implementation. Before: interface rectangle is nearly as tall as the implementation rectangle (shallow). After: interface rectangle is short, implementation rectangle is tall (deep).\n\n### Call-graph collapse\n\nBefore: a tree of function calls rendered as nested boxes. After: the same tree collapsed into one box, with the now-internal calls shown faded inside it.\n\n## Style guidance\n\n- Lean editorial, not corporate-dashboard. Generous whitespace. Serif optional for headings (`font-serif` works well with stone/slate).\n- Colour sparingly: one accent (emerald or indigo) plus red for leakage and amber for warnings.\n- Keep diagrams ~320px tall so before/after sits comfortably side by side without scrolling.\n- Use `text-xs uppercase tracking-wider` for module labels inside diagrams, so they read as schematic, not as UI.\n- The only scripts are the Tailwind CDN and the Mermaid ESM import. The report is otherwise static: no app code, no interactivity beyond Mermaid's own rendering.\n\n## Top recommendation section\n\nOne larger card. Candidate name, one sentence on why, anchor link to its card. That's it.\n\n## Tone\n\nPlain English, concise, but the architectural nouns and verbs come straight from the `/codebase-design` skill. Concision is not an excuse to drift.\n\n**Use exactly:** module, interface, implementation, depth, deep, shallow, seam, adapter, leverage, locality.\n\n**Never substitute:** component, service, unit (for module) · API, signature (for interface) · boundary (for seam) · layer, wrapper (for module, when you mean module).\n\n**Phrasings that fit the style:**\n\n- \"Order intake module is shallow: interface nearly matches the implementation.\"\n- \"Pricing leaks across the seam.\"\n- \"Deepen: one interface, one place to test.\"\n- \"Two adapters justify the seam: HTTP in prod, in-memory in tests.\"\n\n**Wins bullets** name the gain in glossary terms: *\"locality: bugs concentrate in one module\"*, *\"leverage: one interface, N call sites\"*, *\"interface shrinks; implementation absorbs the wrappers\"*. Don't write *\"easier to maintain\"* or *\"cleaner code\"*, because those terms aren't in the glossary and don't earn their place.\n\nNo hedging, no throat-clearing, no \"it's worth noting that…\". If a sentence could be a bullet, make it a bullet. If a bullet could be cut, cut it. If a term isn't in the `/codebase-design` glossary, reach for one that is before inventing a new one.\n",".agents/skills/improve-codebase-architecture/SKILL.md":"---\nname: improve-codebase-architecture\ndescription: Scan a codebase for deepening opportunities, present them as a visual HTML report, then grill through whichever one you pick.\ndisable-model-invocation: true\n---\n\n# Improve Codebase Architecture\n\nSurface architectural friction and propose **deepening opportunities**: refactors that turn shallow modules into deep ones. The aim is testability and AI-navigability.\n\nThis command is _informed_ by the project's domain model and built on a shared design vocabulary:\n\n- Call the Skill tool with \"codebase-design\" for the architecture vocabulary (**module**, **interface**, **depth**, **seam**, **adapter**, **leverage**, **locality**) and its principles (the deletion test, \"the interface is the test surface\", \"one adapter = hypothetical seam, two = real\"). Use these terms exactly in every suggestion, and don't drift into \"component,\" \"service,\" \"API,\" or \"boundary.\"\n- The domain language in `GLOSSARY.md` gives names to good seams; ADRs in `docs/adr/` record decisions this command should not re-litigate.\n\n## Process\n\n### 1. Explore\n\n**Scope before you scan: YAGNI.** Deepening a module pays off by making future changes to it easier, so put extra weight on the parts of the codebase that have recently changed. Decide *where* to look before you look:\n\n- If the user named a direction (a module, a subsystem, a pain point), take it, and skip the inference below.\n- Otherwise, walk back a good stretch of the commit history (`git log --oneline`) to find the codebase's hot spots, the files and areas that keep coming up, and let those paths pull your attention first. If the changes are scattered with no clear hot spot, widen the net.\n\nRead the project's domain glossary (`GLOSSARY.md`) and any ADRs in the area you're touching first.\n\nThen spawn a sub-agent to walk the codebase. Don't follow rigid heuristics; explore organically and note where you experience friction:\n\n- Where does understanding one concept require bouncing between many small modules?\n- Where are modules **shallow**, with an interface nearly as complex as the implementation?\n- Where have pure functions been extracted just for testability, but the real bugs hide in how they're called (no **locality**)?\n- Where do tightly-coupled modules leak across their seams?\n- Which parts of the codebase are untested, or hard to test through their current interface?\n\nApply the **deletion test** to anything you suspect is shallow: would deleting it concentrate complexity, or just move it? A \"yes, concentrates\" is the signal you want.\n\n### 2. Present candidates as an HTML report\n\nWrite a self-contained HTML file to the OS temp directory so nothing lands in the repo. Resolve the temp dir from `$TMPDIR`, falling back to `/tmp` (or `%TEMP%` on Windows), and write to `/architecture-review-.html` so each run gets a fresh file. Open it for the user (`xdg-open ` on Linux, `open ` on macOS, `start ` on Windows) and tell them the absolute path.\n\nThe report uses **Tailwind via CDN** for layout and styling, and **Mermaid via CDN** for diagrams where a graph/flow/sequence reliably communicates the structure. Mix Mermaid with hand-crafted CSS/SVG visuals: use Mermaid when relationships are graph-shaped (call graphs, dependencies, sequences), and hand-built divs/SVG when you want something more editorial (mass diagrams, cross-sections, collapse animations). Each candidate gets a **before/after visualisation**. Be visual.\n\nFor each candidate, render a card with:\n\n- **Files**: which files/modules are involved\n- **Problem**: why the current architecture is causing friction\n- **Solution**: plain English description of what would change\n- **Benefits**: explained in terms of locality and leverage, and how tests would improve\n- **Before / After diagram**: side-by-side, custom-drawn, illustrating the shallowness and the deepening\n- **Recommendation strength**: one of `Strong`, `Worth exploring`, `Speculative`, rendered as a badge\n\nEnd the report with a **Top recommendation** section: which candidate you'd tackle first and why.\n\n**Use GLOSSARY.md vocabulary for the domain, and the `/codebase-design` vocabulary for the architecture.** If `GLOSSARY.md` defines \"Order,\" talk about \"the Order intake module,\" not \"the FooBarHandler,\" and not \"the Order service.\"\n\n**ADR conflicts**: if a candidate contradicts an existing ADR, only surface it when the friction is real enough to warrant revisiting the ADR. Mark it clearly in the card (e.g. a warning callout: _\"contradicts ADR-0007, but worth reopening because…\"_). Don't list every theoretical refactor an ADR forbids.\n\nSee [HTML-REPORT.md](HTML-REPORT.md) for the full HTML scaffold, diagram patterns, and styling guidance.\n\nDo NOT propose interfaces yet. After the file is written, ask the user: \"Which of these would you like to explore?\"\n\n### 3. Grilling loop\n\nOnce the user picks a candidate, call the Skill tool with \"grilling\" to walk the decision tree with them: constraints, dependencies, the shape of the deepened module, what sits behind the seam, what tests survive.\n\nSide effects happen inline as decisions crystallize; call the Skill tool with \"domain-modeling\" to keep the domain model current as you go:\n\n- **Naming a deepened module after a concept not in `GLOSSARY.md`?** Add the term to `GLOSSARY.md`. Create the file lazily if it doesn't exist.\n- **Sharpening a fuzzy term during the conversation?** Update `GLOSSARY.md` right there.\n- **User rejects the candidate with a load-bearing reason?** Offer an ADR, framed as: _\"Want me to record this as an ADR so future architecture reviews don't re-suggest it?\"_ Only offer when the reason would actually be needed by a future explorer to avoid re-suggesting the same thing; skip ephemeral reasons (\"not worth it right now\") and self-evident ones.\n- **Want to explore alternative interfaces for the deepened module?** Call the Skill tool with \"codebase-design\" and use its design-it-twice parallel sub-agent pattern.\n",".agents/skills/improve-codebase-architecture/agents/openai.yaml":"interface:\n display_name: \"Improve Codebase Architecture\"\n short_description: \"Find and grill architecture improvements\"\npolicy:\n allow_implicit_invocation: false\n",".agents/skills/pr/CREDITS.md":"# Credits\n\nThe **Summary** section's menu of visuals (pseudocode, call trees, component trees, file trees, Mermaid, diffs) and its placement guidance come from [Dex Horthy](https://github.com/dexhorthy)'s [`show-me`](https://github.com/humanlayer/humanlayer) skill, reproduced almost word for word and aimed at a diff instead of a live conversation. `pr` does not depend on `show-me` as a skill (it isn't part of this repo, and a hard dependency would break standalone installs), so the content is copied in rather than pointed at; this file is the attribution a dependency would otherwise have carried.\n",".agents/skills/pr/SKILL.md":"---\nname: pr\ndescription: \"Use when writing a PR body.\"\nmetadata:\n credits:\n skill: show-me\n author: Dex Horthy\n organisation: Humanlayer\n url: \"https://github.com/humanlayer/skills/blob/main/plugins/show-me/skills/show-me/SKILL.md\"\n---\n\nUse this template for writing the PR body:\n\n```markdown\n## Summary\n\n\n\n## Evidence\n\n- **Before:** \n **After:** \n\n## Merge Danger\n\n**Door:** \n\n\n\n**Blast Radius:** \n\n\n```\n\n## Sections\n\nSkip all preambles and keep prose brief. Use the user's domain language from `GLOSSARY.md`.\n\n### Summary\n\nPick the smallest view that makes the key point clear.\n\n- Show logic or an algorithm as pseudocode:\n\n```text\non(save)\n if content is unchanged\n return cached result\n write new content\n return fresh result\n```\n\n- Show runtime control flow as a call tree:\n\n```text\nsubmitForm\n createSession\n persistPrompt\n launchAgent\n navigateToSession\n```\n\n- Show UI structure as a component tree, including state and module boundaries that matter:\n\n```text\n (apps/example/src/routes/session.tsx)\n useSessionEvents()\n \n (packages/ui)\n```\n\n- Show file responsibility or a broad refactor as a shallow file tree:\n\n```text\nsrc/\n├── commands/ # parses user actions\n├── sessions/ # owns session state\n└── transport/ # sends API requests\n```\n\n- Show component interaction, control flow, or data flow with Mermaid:\n\n```mermaid\nsequenceDiagram\n participant User\n participant UI\n participant Daemon\n User->>UI: choose command\n UI->>Daemon: send expanded prompt\n Daemon-->>UI: stream result\n```\n\n- Use `diff` when the point is what changes and the surrounding shape already exists. Match the diff shape to the topic.\n\nFor a component change:\n\n```diff\n \n useSessionEvents()\n \n+ \n \n+ \n```\n\nFor a file-layout change:\n\n```diff\n src/\n ├── commands/\n+│ └── show-me.ts # expands the slash command\n ├── sessions/\n-└── transport.ts\n+└── transport/\n+ ├── client.ts\n+ └── stream.ts\n```\n\nFor a call-tree or call-stack change:\n\n```diff\n submitForm\n createSession\n persistPrompt\n+ expandSkillMention\n launchAgent\n- navigateToSession\n+ navigateToSession\n+ subscribeToEvents\n```\n\nFor a state or control-flow change:\n\n```diff\n on(save)\n- write content\n+ if content is unchanged\n+ return cached result\n+ write new content\n+ invalidate cache\n```\n\n- Show the whole block when most of it is new, when omitted context would hide ownership or order, or when the user needs a copyable target shape:\n\n```ts\nfunction expandSkill(command: string): string {\n const skillName = command.slice(1);\n return `use the ${skillName} skill`;\n}\n```\n\n#### Guidance\n\nPlace each visual next to the short text it supports. Keep only the calls, files, props, states, and boundaries needed to answer the user's current question or the options to resolve the current discussion point.\n\nYou may use one of these, you may use several, it is unlikely you will use all of them. Use your judgement and don't overwhelm the user.\n\n### Evidence\n\nConcrete evidence that the change works. Show a before and after.\n\nScreenshots are S-tier - when the environment is set up for it and the change is visual.\n\nExecution-based evidence is A-tier. Test results, console output. Show the exact test that now fails and passes, using pseudocode.\n\n### Merge Danger\n\nDescribe whether it's a one-way or two-way door. You can walk back through two-way doors, but not one-way doors. A PR that is cheap to roll back is lower risk. Changes that involve destructive actions or hard-to-reverse decisions are one-way doors.\n\nThe blast radius is the potential impact or scope of the changes introduced by this PR. Consider all possibilities. Examples are layout shift, breakages for consumers, mobile responsiveness, etc.\n",".agents/skills/pr/agents/openai.yaml":"interface:\n display_name: \"PR\"\n short_description: \"Write a PR body that's fast to review\"\n",".agents/skills/prototype/LOGIC.md":"# Logic Prototype\n\nA single, self-contained HTML file (a **shareable demo**) that lets anyone drive a state model by clicking buttons. Use this when the question is about **business logic, state transitions, or data shape**: the kind of thing that looks reasonable on paper but only feels wrong once you push it through real cases.\n\nBecause it's one file with nothing to install, you can hand it to a non-developer (a designer, a PM, a domain expert) and let them feel the model for themselves. So it speaks their language, not the code's.\n\n## When this is the right shape\n\n- \"I'm not sure if this state machine handles the edge case where X then Y.\"\n- \"Does this data model actually let me represent the case where...\"\n- \"I want to feel out what the API should look like before writing it.\"\n- Anything where someone wants to **press buttons and watch state change**.\n\nIf the question is \"what should this look like,\" this is the wrong branch. Use [UI.md](UI.md).\n\n## Process\n\n### 1. State the question\n\nBefore writing code, write down what state model and what question you're prototyping. One paragraph, at the top of the demo (in a visible intro, not just a comment). A logic prototype that answers the wrong question is pure waste, so make the question explicit so it can be checked later, whether the user is watching now or returning to it AFK.\n\n### 2. Isolate the logic in a portable module\n\nPut the actual logic (the bit that's answering the question) in a single `\n \n \n \n \n
\n
...
\n
...
\n
...
\n
\n \n\n```\n\n## Header\n\nRepo name, date, and a compact legend: solid box = module, dashed line = seam, red arrow = leakage, thick dark box = deep module. No introduction paragraph. Straight into the candidates.\n\n## Candidate card\n\nThe diagrams carry the weight. Prose is sparse, plain, and uses the glossary terms (from the `/codebase-design` skill) without ceremony.\n\nEach candidate is one `
`:\n\n- **Title**: short, names the deepening (e.g. \"Collapse the Order intake pipeline\").\n- **Badge row**: recommendation strength (`Strong` = emerald, `Worth exploring` = amber, `Speculative` = slate), plus a tag for the dependency category (`in-process`, `local-substitutable`, `ports & adapters`, `mock`).\n- **Files**: monospaced list, `font-mono text-sm`.\n- **Before / After diagram**: the centrepiece. Two columns, side by side. See patterns below.\n- **Problem**: one sentence. What hurts.\n- **Solution**: one sentence. What changes.\n- **Wins**: bullets, ≤6 words each. e.g. \"Tests hit one interface\", \"Pricing logic stops leaking\", \"Delete 4 shallow wrappers\".\n- **ADR callout** (if applicable): one line in an amber-tinted box.\n\nNo paragraphs of explanation. If the diagram needs a paragraph to be understood, redraw the diagram.\n\n## Diagram patterns\n\nPick the pattern that fits the candidate. Mix them. Don't make every diagram look the same. Variety is part of the point.\n\n### Mermaid graph (the workhorse for dependencies / call flow)\n\nUse a Mermaid `flowchart` or `graph` when the point is \"X calls Y calls Z, and look at the mess.\" Wrap it in a Tailwind-styled card so it doesn't feel parachuted in. Style with classDef to colour leakage edges red and the deep module dark. Sequence diagrams work well for \"before: 6 round-trips; after: 1.\"\n\n```html\n
\n 
\n    flowchart LR\n      A[OrderHandler] --> B[OrderValidator]\n      B --> C[OrderRepo]\n      C -.leak.-> D[PricingClient]\n      classDef leak stroke:#dc2626,stroke-width:2px;\n      class C,D leak\n
\n
\n```\n\n### Hand-built boxes-and-arrows (when Mermaid's layout fights you)\n\nModules as `
`s with borders and labels. Arrows as inline SVG `` or `` elements positioned absolutely over a relative container. Reach for this when you want the \"after\" diagram to feel like one thick-bordered deep module with greyed-out internals, since Mermaid won't render that with the right weight.\n\n### Cross-section (good for layered shallowness)\n\nStack horizontal bands (`h-12 border-l-4`) to show layers a call passes through. Before: 6 thin layers each doing nothing. After: 1 thick band labelled with the consolidated responsibility.\n\n### Mass diagram (good for \"interface as wide as implementation\")\n\nTwo rectangles per module: one for interface surface area, one for implementation. Before: interface rectangle is nearly as tall as the implementation rectangle (shallow). After: interface rectangle is short, implementation rectangle is tall (deep).\n\n### Call-graph collapse\n\nBefore: a tree of function calls rendered as nested boxes. After: the same tree collapsed into one box, with the now-internal calls shown faded inside it.\n\n## Style guidance\n\n- Lean editorial, not corporate-dashboard. Generous whitespace. Serif optional for headings (`font-serif` works well with stone/slate).\n- Colour sparingly: one accent (emerald or indigo) plus red for leakage and amber for warnings.\n- Keep diagrams ~320px tall so before/after sits comfortably side by side without scrolling.\n- Use `text-xs uppercase tracking-wider` for module labels inside diagrams, so they read as schematic, not as UI.\n- The only scripts are the Tailwind CDN and the Mermaid ESM import. The report is otherwise static: no app code, no interactivity beyond Mermaid's own rendering.\n\n## Top recommendation section\n\nOne larger card. Candidate name, one sentence on why, anchor link to its card. That's it.\n\n## Tone\n\nPlain English, concise, but the architectural nouns and verbs come straight from the `/codebase-design` skill. Concision is not an excuse to drift.\n\n**Use exactly:** module, interface, implementation, depth, deep, shallow, seam, adapter, leverage, locality.\n\n**Never substitute:** component, service, unit (for module) · API, signature (for interface) · boundary (for seam) · layer, wrapper (for module, when you mean module).\n\n**Phrasings that fit the style:**\n\n- \"Order intake module is shallow: interface nearly matches the implementation.\"\n- \"Pricing leaks across the seam.\"\n- \"Deepen: one interface, one place to test.\"\n- \"Two adapters justify the seam: HTTP in prod, in-memory in tests.\"\n\n**Wins bullets** name the gain in glossary terms: *\"locality: bugs concentrate in one module\"*, *\"leverage: one interface, N call sites\"*, *\"interface shrinks; implementation absorbs the wrappers\"*. Don't write *\"easier to maintain\"* or *\"cleaner code\"*, because those terms aren't in the glossary and don't earn their place.\n\nNo hedging, no throat-clearing, no \"it's worth noting that…\". If a sentence could be a bullet, make it a bullet. If a bullet could be cut, cut it. If a term isn't in the `/codebase-design` glossary, reach for one that is before inventing a new one.\n",".agents/skills/improve-codebase-architecture/SKILL.md":"---\nname: improve-codebase-architecture\ndescription: Scan a codebase for deepening opportunities, present them as a visual HTML report, then grill through whichever one you pick.\ndisable-model-invocation: true\n---\n\n# Improve Codebase Architecture\n\nSurface architectural friction and propose **deepening opportunities**: refactors that turn shallow modules into deep ones. The aim is testability and AI-navigability.\n\nThis command is _informed_ by the project's domain model and built on a shared design vocabulary:\n\n- Call the Skill tool with \"codebase-design\" for the architecture vocabulary (**module**, **interface**, **depth**, **seam**, **adapter**, **leverage**, **locality**) and its principles (the deletion test, \"the interface is the test surface\", \"one adapter = hypothetical seam, two = real\"). Use these terms exactly in every suggestion, and don't drift into \"component,\" \"service,\" \"API,\" or \"boundary.\"\n- The domain language in `GLOSSARY.md` gives names to good seams; ADRs in `docs/adr/` record decisions this command should not re-litigate.\n\n## Process\n\n### 1. Explore\n\n**Scope before you scan: YAGNI.** Deepening a module pays off by making future changes to it easier, so put extra weight on the parts of the codebase that have recently changed. Decide *where* to look before you look:\n\n- If the user named a direction (a module, a subsystem, a pain point), take it, and skip the inference below.\n- Otherwise, walk back a good stretch of the commit history (`git log --oneline`) to find the codebase's hot spots, the files and areas that keep coming up, and let those paths pull your attention first. If the changes are scattered with no clear hot spot, widen the net.\n\nRead the project's domain glossary (`GLOSSARY.md`) and any ADRs in the area you're touching first.\n\nThen spawn a sub-agent to walk the codebase. Don't follow rigid heuristics; explore organically and note where you experience friction:\n\n- Where does understanding one concept require bouncing between many small modules?\n- Where are modules **shallow**, with an interface nearly as complex as the implementation?\n- Where have pure functions been extracted just for testability, but the real bugs hide in how they're called (no **locality**)?\n- Where do tightly-coupled modules leak across their seams?\n- Which parts of the codebase are untested, or hard to test through their current interface?\n\nApply the **deletion test** to anything you suspect is shallow: would deleting it concentrate complexity, or just move it? A \"yes, concentrates\" is the signal you want.\n\n### 2. Present candidates as an HTML report\n\nWrite a self-contained HTML file to the OS temp directory so nothing lands in the repo. Resolve the temp dir from `$TMPDIR`, falling back to `/tmp` (or `%TEMP%` on Windows), and write to `/architecture-review-.html` so each run gets a fresh file. Open it for the user (`xdg-open ` on Linux, `open ` on macOS, `start ` on Windows) and tell them the absolute path.\n\nThe report uses **Tailwind via CDN** for layout and styling, and **Mermaid via CDN** for diagrams where a graph/flow/sequence reliably communicates the structure. Mix Mermaid with hand-crafted CSS/SVG visuals: use Mermaid when relationships are graph-shaped (call graphs, dependencies, sequences), and hand-built divs/SVG when you want something more editorial (mass diagrams, cross-sections, collapse animations). Each candidate gets a **before/after visualisation**. Be visual.\n\nFor each candidate, render a card with:\n\n- **Files**: which files/modules are involved\n- **Problem**: why the current architecture is causing friction\n- **Solution**: plain English description of what would change\n- **Benefits**: explained in terms of locality and leverage, and how tests would improve\n- **Before / After diagram**: side-by-side, custom-drawn, illustrating the shallowness and the deepening\n- **Recommendation strength**: one of `Strong`, `Worth exploring`, `Speculative`, rendered as a badge\n\nEnd the report with a **Top recommendation** section: which candidate you'd tackle first and why.\n\n**Use GLOSSARY.md vocabulary for the domain, and the `/codebase-design` vocabulary for the architecture.** If `GLOSSARY.md` defines \"Order,\" talk about \"the Order intake module,\" not \"the FooBarHandler,\" and not \"the Order service.\"\n\n**ADR conflicts**: if a candidate contradicts an existing ADR, only surface it when the friction is real enough to warrant revisiting the ADR. Mark it clearly in the card (e.g. a warning callout: _\"contradicts ADR-0007, but worth reopening because…\"_). Don't list every theoretical refactor an ADR forbids.\n\nSee [HTML-REPORT.md](HTML-REPORT.md) for the full HTML scaffold, diagram patterns, and styling guidance.\n\nDo NOT propose interfaces yet. After the file is written, ask the user: \"Which of these would you like to explore?\"\n\n### 3. Grilling loop\n\nOnce the user picks a candidate, call the Skill tool with \"grilling\" to walk the decision tree with them: constraints, dependencies, the shape of the deepened module, what sits behind the seam, what tests survive.\n\nSide effects happen inline as decisions crystallize; call the Skill tool with \"domain-modeling\" to keep the domain model current as you go:\n\n- **Naming a deepened module after a concept not in `GLOSSARY.md`?** Add the term to `GLOSSARY.md`. Create the file lazily if it doesn't exist.\n- **Sharpening a fuzzy term during the conversation?** Update `GLOSSARY.md` right there.\n- **User rejects the candidate with a load-bearing reason?** Offer an ADR, framed as: _\"Want me to record this as an ADR so future architecture reviews don't re-suggest it?\"_ Only offer when the reason would actually be needed by a future explorer to avoid re-suggesting the same thing; skip ephemeral reasons (\"not worth it right now\") and self-evident ones.\n- **Want to explore alternative interfaces for the deepened module?** Call the Skill tool with \"codebase-design\" and use its design-it-twice parallel sub-agent pattern.\n",".agents/skills/improve-codebase-architecture/agents/openai.yaml":"interface:\n display_name: \"Improve Codebase Architecture\"\n short_description: \"Find and grill architecture improvements\"\npolicy:\n allow_implicit_invocation: false\n",".agents/skills/pr/CREDITS.md":"# Credits\n\nThe **Summary** section's menu of visuals (pseudocode, call trees, component trees, file trees, Mermaid, diffs) and its placement guidance come from [Dex Horthy](https://github.com/dexhorthy)'s [`show-me`](https://github.com/humanlayer/humanlayer) skill, reproduced almost word for word and aimed at a diff instead of a live conversation. `pr` does not depend on `show-me` as a skill (it isn't part of this repo, and a hard dependency would break standalone installs), so the content is copied in rather than pointed at; this file is the attribution a dependency would otherwise have carried.\n",".agents/skills/pr/SKILL.md":"---\nname: pr\ndescription: \"Use when writing a PR body.\"\nmetadata:\n credits:\n skill: show-me\n author: Dex Horthy\n organisation: Humanlayer\n url: \"https://github.com/humanlayer/skills/blob/main/plugins/show-me/skills/show-me/SKILL.md\"\n---\n\nUse this template for writing the PR body:\n\n```markdown\n## Summary\n\n\n\n## Evidence\n\n- **Before:** \n **After:** \n\n## Merge Danger\n\n**Door:** \n\n\n\n**Blast Radius:** \n\n\n```\n\n## Sections\n\nSkip all preambles and keep prose brief. Use the user's domain language from `GLOSSARY.md`.\n\n### Summary\n\nPick the smallest view that makes the key point clear.\n\n- Show logic or an algorithm as pseudocode:\n\n```text\non(save)\n if content is unchanged\n return cached result\n write new content\n return fresh result\n```\n\n- Show runtime control flow as a call tree:\n\n```text\nsubmitForm\n createSession\n persistPrompt\n launchAgent\n navigateToSession\n```\n\n- Show UI structure as a component tree, including state and module boundaries that matter:\n\n```text\n (apps/example/src/routes/session.tsx)\n useSessionEvents()\n \n (packages/ui)\n```\n\n- Show file responsibility or a broad refactor as a shallow file tree:\n\n```text\nsrc/\n├── commands/ # parses user actions\n├── sessions/ # owns session state\n└── transport/ # sends API requests\n```\n\n- Show component interaction, control flow, or data flow with Mermaid:\n\n```mermaid\nsequenceDiagram\n participant User\n participant UI\n participant Daemon\n User->>UI: choose command\n UI->>Daemon: send expanded prompt\n Daemon-->>UI: stream result\n```\n\n- Use `diff` when the point is what changes and the surrounding shape already exists. Match the diff shape to the topic.\n\nFor a component change:\n\n```diff\n \n useSessionEvents()\n \n+ \n \n+ \n```\n\nFor a file-layout change:\n\n```diff\n src/\n ├── commands/\n+│ └── show-me.ts # expands the slash command\n ├── sessions/\n-└── transport.ts\n+└── transport/\n+ ├── client.ts\n+ └── stream.ts\n```\n\nFor a call-tree or call-stack change:\n\n```diff\n submitForm\n createSession\n persistPrompt\n+ expandSkillMention\n launchAgent\n- navigateToSession\n+ navigateToSession\n+ subscribeToEvents\n```\n\nFor a state or control-flow change:\n\n```diff\n on(save)\n- write content\n+ if content is unchanged\n+ return cached result\n+ write new content\n+ invalidate cache\n```\n\n- Show the whole block when most of it is new, when omitted context would hide ownership or order, or when the user needs a copyable target shape:\n\n```ts\nfunction expandSkill(command: string): string {\n const skillName = command.slice(1);\n return `use the ${skillName} skill`;\n}\n```\n\n#### Guidance\n\nPlace each visual next to the short text it supports. Keep only the calls, files, props, states, and boundaries needed to answer the user's current question or the options to resolve the current discussion point.\n\nYou may use one of these, you may use several, it is unlikely you will use all of them. Use your judgement and don't overwhelm the user.\n\n### Evidence\n\nConcrete evidence that the change works. Show a before and after.\n\nScreenshots are S-tier - when the environment is set up for it and the change is visual.\n\nExecution-based evidence is A-tier. Test results, console output. Show the exact test that now fails and passes, using pseudocode.\n\n### Merge Danger\n\nDescribe whether it's a one-way or two-way door. You can walk back through two-way doors, but not one-way doors. A PR that is cheap to roll back is lower risk. Changes that involve destructive actions or hard-to-reverse decisions are one-way doors.\n\nThe blast radius is the potential impact or scope of the changes introduced by this PR. Consider all possibilities. Examples are layout shift, breakages for consumers, mobile responsiveness, etc.\n",".agents/skills/pr/agents/openai.yaml":"interface:\n display_name: \"PR\"\n short_description: \"Write a PR body that's fast to review\"\n",".agents/skills/prototype/LOGIC.md":"# Logic Prototype\n\nA single, self-contained HTML file (a **shareable demo**) that lets anyone drive a state model by clicking buttons. Use this when the question is about **business logic, state transitions, or data shape**: the kind of thing that looks reasonable on paper but only feels wrong once you push it through real cases.\n\nBecause it's one file with nothing to install, you can hand it to a non-developer (a designer, a PM, a domain expert) and let them feel the model for themselves. So it speaks their language, not the code's.\n\n## When this is the right shape\n\n- \"I'm not sure if this state machine handles the edge case where X then Y.\"\n- \"Does this data model actually let me represent the case where...\"\n- \"I want to feel out what the API should look like before writing it.\"\n- Anything where someone wants to **press buttons and watch state change**.\n\nIf the question is \"what should this look like,\" this is the wrong branch. Use [UI.md](UI.md).\n\n## Process\n\n### 1. State the question\n\nBefore writing code, write down what state model and what question you're prototyping. One paragraph, at the top of the demo (in a visible intro, not just a comment). A logic prototype that answers the wrong question is pure waste, so make the question explicit so it can be checked later, whether the user is watching now or returning to it AFK.\n\n### 2. Isolate the logic in a portable module\n\nPut the actual logic (the bit that's answering the question) in a single `\n \n \n \n \n
\n
...
\n
...
\n
...
\n
\n \n\n```\n\n## Header\n\nRepo name, date, and a compact legend: solid box = module, dashed line = seam, red arrow = leakage, thick dark box = deep module. No introduction paragraph. Straight into the candidates.\n\n## Candidate card\n\nThe diagrams carry the weight. Prose is sparse, plain, and uses the glossary terms (from the `/codebase-design` skill) without ceremony.\n\nEach candidate is one `
`:\n\n- **Title**: short, names the deepening (e.g. \"Collapse the Order intake pipeline\").\n- **Badge row**: recommendation strength (`Strong` = emerald, `Worth exploring` = amber, `Speculative` = slate), plus a tag for the dependency category (`in-process`, `local-substitutable`, `ports & adapters`, `mock`).\n- **Files**: monospaced list, `font-mono text-sm`.\n- **Before / After diagram**: the centrepiece. Two columns, side by side. See patterns below.\n- **Problem**: one sentence. What hurts.\n- **Solution**: one sentence. What changes.\n- **Wins**: bullets, ≤6 words each. e.g. \"Tests hit one interface\", \"Pricing logic stops leaking\", \"Delete 4 shallow wrappers\".\n- **ADR callout** (if applicable): one line in an amber-tinted box.\n\nNo paragraphs of explanation. If the diagram needs a paragraph to be understood, redraw the diagram.\n\n## Diagram patterns\n\nPick the pattern that fits the candidate. Mix them. Don't make every diagram look the same. Variety is part of the point.\n\n### Mermaid graph (the workhorse for dependencies / call flow)\n\nUse a Mermaid `flowchart` or `graph` when the point is \"X calls Y calls Z, and look at the mess.\" Wrap it in a Tailwind-styled card so it doesn't feel parachuted in. Style with classDef to colour leakage edges red and the deep module dark. Sequence diagrams work well for \"before: 6 round-trips; after: 1.\"\n\n```html\n
\n 
\n    flowchart LR\n      A[OrderHandler] --> B[OrderValidator]\n      B --> C[OrderRepo]\n      C -.leak.-> D[PricingClient]\n      classDef leak stroke:#dc2626,stroke-width:2px;\n      class C,D leak\n
\n
\n```\n\n### Hand-built boxes-and-arrows (when Mermaid's layout fights you)\n\nModules as `
`s with borders and labels. Arrows as inline SVG `` or `` elements positioned absolutely over a relative container. Reach for this when you want the \"after\" diagram to feel like one thick-bordered deep module with greyed-out internals, since Mermaid won't render that with the right weight.\n\n### Cross-section (good for layered shallowness)\n\nStack horizontal bands (`h-12 border-l-4`) to show layers a call passes through. Before: 6 thin layers each doing nothing. After: 1 thick band labelled with the consolidated responsibility.\n\n### Mass diagram (good for \"interface as wide as implementation\")\n\nTwo rectangles per module: one for interface surface area, one for implementation. Before: interface rectangle is nearly as tall as the implementation rectangle (shallow). After: interface rectangle is short, implementation rectangle is tall (deep).\n\n### Call-graph collapse\n\nBefore: a tree of function calls rendered as nested boxes. After: the same tree collapsed into one box, with the now-internal calls shown faded inside it.\n\n## Style guidance\n\n- Lean editorial, not corporate-dashboard. Generous whitespace. Serif optional for headings (`font-serif` works well with stone/slate).\n- Colour sparingly: one accent (emerald or indigo) plus red for leakage and amber for warnings.\n- Keep diagrams ~320px tall so before/after sits comfortably side by side without scrolling.\n- Use `text-xs uppercase tracking-wider` for module labels inside diagrams, so they read as schematic, not as UI.\n- The only scripts are the Tailwind CDN and the Mermaid ESM import. The report is otherwise static: no app code, no interactivity beyond Mermaid's own rendering.\n\n## Top recommendation section\n\nOne larger card. Candidate name, one sentence on why, anchor link to its card. That's it.\n\n## Tone\n\nPlain English, concise, but the architectural nouns and verbs come straight from the `/codebase-design` skill. Concision is not an excuse to drift.\n\n**Use exactly:** module, interface, implementation, depth, deep, shallow, seam, adapter, leverage, locality.\n\n**Never substitute:** component, service, unit (for module) · API, signature (for interface) · boundary (for seam) · layer, wrapper (for module, when you mean module).\n\n**Phrasings that fit the style:**\n\n- \"Order intake module is shallow: interface nearly matches the implementation.\"\n- \"Pricing leaks across the seam.\"\n- \"Deepen: one interface, one place to test.\"\n- \"Two adapters justify the seam: HTTP in prod, in-memory in tests.\"\n\n**Wins bullets** name the gain in glossary terms: *\"locality: bugs concentrate in one module\"*, *\"leverage: one interface, N call sites\"*, *\"interface shrinks; implementation absorbs the wrappers\"*. Don't write *\"easier to maintain\"* or *\"cleaner code\"*, because those terms aren't in the glossary and don't earn their place.\n\nNo hedging, no throat-clearing, no \"it's worth noting that…\". If a sentence could be a bullet, make it a bullet. If a bullet could be cut, cut it. If a term isn't in the `/codebase-design` glossary, reach for one that is before inventing a new one.\n","isInternal":false,"tokens":1610,"sizeBytes":6641},{"name":"SKILL.md","path":".agents/skills/improve-codebase-architecture/SKILL.md","rawUrl":"https://raw.githubusercontent.com/withcoral/coral/HEAD/.agents/skills/improve-codebase-architecture/SKILL.md","title":"improve-codebase-architecture","category":"anthropic-skill","format":"markdown","content":"---\nname: improve-codebase-architecture\ndescription: Scan a codebase for deepening opportunities, present them as a visual HTML report, then grill through whichever one you pick.\ndisable-model-invocation: true\n---\n\n# Improve Codebase Architecture\n\nSurface architectural friction and propose **deepening opportunities**: refactors that turn shallow modules into deep ones. The aim is testability and AI-navigability.\n\nThis command is _informed_ by the project's domain model and built on a shared design vocabulary:\n\n- Call the Skill tool with \"codebase-design\" for the architecture vocabulary (**module**, **interface**, **depth**, **seam**, **adapter**, **leverage**, **locality**) and its principles (the deletion test, \"the interface is the test surface\", \"one adapter = hypothetical seam, two = real\"). Use these terms exactly in every suggestion, and don't drift into \"component,\" \"service,\" \"API,\" or \"boundary.\"\n- The domain language in `GLOSSARY.md` gives names to good seams; ADRs in `docs/adr/` record decisions this command should not re-litigate.\n\n## Process\n\n### 1. Explore\n\n**Scope before you scan: YAGNI.** Deepening a module pays off by making future changes to it easier, so put extra weight on the parts of the codebase that have recently changed. Decide *where* to look before you look:\n\n- If the user named a direction (a module, a subsystem, a pain point), take it, and skip the inference below.\n- Otherwise, walk back a good stretch of the commit history (`git log --oneline`) to find the codebase's hot spots, the files and areas that keep coming up, and let those paths pull your attention first. If the changes are scattered with no clear hot spot, widen the net.\n\nRead the project's domain glossary (`GLOSSARY.md`) and any ADRs in the area you're touching first.\n\nThen spawn a sub-agent to walk the codebase. Don't follow rigid heuristics; explore organically and note where you experience friction:\n\n- Where does understanding one concept require bouncing between many small modules?\n- Where are modules **shallow**, with an interface nearly as complex as the implementation?\n- Where have pure functions been extracted just for testability, but the real bugs hide in how they're called (no **locality**)?\n- Where do tightly-coupled modules leak across their seams?\n- Which parts of the codebase are untested, or hard to test through their current interface?\n\nApply the **deletion test** to anything you suspect is shallow: would deleting it concentrate complexity, or just move it? A \"yes, concentrates\" is the signal you want.\n\n### 2. Present candidates as an HTML report\n\nWrite a self-contained HTML file to the OS temp directory so nothing lands in the repo. Resolve the temp dir from `$TMPDIR`, falling back to `/tmp` (or `%TEMP%` on Windows), and write to `/architecture-review-.html` so each run gets a fresh file. Open it for the user (`xdg-open ` on Linux, `open ` on macOS, `start ` on Windows) and tell them the absolute path.\n\nThe report uses **Tailwind via CDN** for layout and styling, and **Mermaid via CDN** for diagrams where a graph/flow/sequence reliably communicates the structure. Mix Mermaid with hand-crafted CSS/SVG visuals: use Mermaid when relationships are graph-shaped (call graphs, dependencies, sequences), and hand-built divs/SVG when you want something more editorial (mass diagrams, cross-sections, collapse animations). Each candidate gets a **before/after visualisation**. Be visual.\n\nFor each candidate, render a card with:\n\n- **Files**: which files/modules are involved\n- **Problem**: why the current architecture is causing friction\n- **Solution**: plain English description of what would change\n- **Benefits**: explained in terms of locality and leverage, and how tests would improve\n- **Before / After diagram**: side-by-side, custom-drawn, illustrating the shallowness and the deepening\n- **Recommendation strength**: one of `Strong`, `Worth exploring`, `Speculative`, rendered as a badge\n\nEnd the report with a **Top recommendation** section: which candidate you'd tackle first and why.\n\n**Use GLOSSARY.md vocabulary for the domain, and the `/codebase-design` vocabulary for the architecture.** If `GLOSSARY.md` defines \"Order,\" talk about \"the Order intake module,\" not \"the FooBarHandler,\" and not \"the Order service.\"\n\n**ADR conflicts**: if a candidate contradicts an existing ADR, only surface it when the friction is real enough to warrant revisiting the ADR. Mark it clearly in the card (e.g. a warning callout: _\"contradicts ADR-0007, but worth reopening because…\"_). Don't list every theoretical refactor an ADR forbids.\n\nSee [HTML-REPORT.md](HTML-REPORT.md) for the full HTML scaffold, diagram patterns, and styling guidance.\n\nDo NOT propose interfaces yet. After the file is written, ask the user: \"Which of these would you like to explore?\"\n\n### 3. Grilling loop\n\nOnce the user picks a candidate, call the Skill tool with \"grilling\" to walk the decision tree with them: constraints, dependencies, the shape of the deepened module, what sits behind the seam, what tests survive.\n\nSide effects happen inline as decisions crystallize; call the Skill tool with \"domain-modeling\" to keep the domain model current as you go:\n\n- **Naming a deepened module after a concept not in `GLOSSARY.md`?** Add the term to `GLOSSARY.md`. Create the file lazily if it doesn't exist.\n- **Sharpening a fuzzy term during the conversation?** Update `GLOSSARY.md` right there.\n- **User rejects the candidate with a load-bearing reason?** Offer an ADR, framed as: _\"Want me to record this as an ADR so future architecture reviews don't re-suggest it?\"_ Only offer when the reason would actually be needed by a future explorer to avoid re-suggesting the same thing; skip ephemeral reasons (\"not worth it right now\") and self-evident ones.\n- **Want to explore alternative interfaces for the deepened module?** Call the Skill tool with \"codebase-design\" and use its design-it-twice parallel sub-agent pattern.\n","frontmatter":{"name":"improve-codebase-architecture","description":"Scan a codebase for deepening opportunities, present them as a visual HTML report, then grill through whichever one you pick.","disable-model-invocation":true,"disableModelInvocation":true},"isInternal":false,"tokens":1358,"sizeBytes":6000},{"name":"openai.yaml","path":".agents/skills/pr/agents/openai.yaml","rawUrl":"https://raw.githubusercontent.com/withcoral/coral/HEAD/.agents/skills/pr/agents/openai.yaml","title":"Agents Skill","category":"anthropic-skill","format":"yaml","content":"interface:\n display_name: \"PR\"\n short_description: \"Write a PR body that's fast to review\"\n","isInternal":false,"tokens":24,"sizeBytes":93},{"name":"CREDITS.md","path":".agents/skills/pr/CREDITS.md","rawUrl":"https://raw.githubusercontent.com/withcoral/coral/HEAD/.agents/skills/pr/CREDITS.md","title":"Pr Skill","category":"anthropic-skill","format":"markdown","content":"# Credits\n\nThe **Summary** section's menu of visuals (pseudocode, call trees, component trees, file trees, Mermaid, diffs) and its placement guidance come from [Dex Horthy](https://github.com/dexhorthy)'s [`show-me`](https://github.com/humanlayer/humanlayer) skill, reproduced almost word for word and aimed at a diff instead of a live conversation. `pr` does not depend on `show-me` as a skill (it isn't part of this repo, and a hard dependency would break standalone installs), so the content is copied in rather than pointed at; this file is the attribution a dependency would otherwise have carried.\n","isInternal":false,"tokens":141,"sizeBytes":604},{"name":"SKILL.md","path":".agents/skills/pr/SKILL.md","rawUrl":"https://raw.githubusercontent.com/withcoral/coral/HEAD/.agents/skills/pr/SKILL.md","title":"pr","category":"anthropic-skill","format":"markdown","content":"---\nname: pr\ndescription: \"Use when writing a PR body.\"\nmetadata:\n credits:\n skill: show-me\n author: Dex Horthy\n organisation: Humanlayer\n url: \"https://github.com/humanlayer/skills/blob/main/plugins/show-me/skills/show-me/SKILL.md\"\n---\n\nUse this template for writing the PR body:\n\n```markdown\n## Summary\n\n\n\n## Evidence\n\n- **Before:** \n **After:** \n\n## Merge Danger\n\n**Door:** \n\n\n\n**Blast Radius:** \n\n\n```\n\n## Sections\n\nSkip all preambles and keep prose brief. Use the user's domain language from `GLOSSARY.md`.\n\n### Summary\n\nPick the smallest view that makes the key point clear.\n\n- Show logic or an algorithm as pseudocode:\n\n```text\non(save)\n if content is unchanged\n return cached result\n write new content\n return fresh result\n```\n\n- Show runtime control flow as a call tree:\n\n```text\nsubmitForm\n createSession\n persistPrompt\n launchAgent\n navigateToSession\n```\n\n- Show UI structure as a component tree, including state and module boundaries that matter:\n\n```text\n (apps/example/src/routes/session.tsx)\n useSessionEvents()\n \n (packages/ui)\n```\n\n- Show file responsibility or a broad refactor as a shallow file tree:\n\n```text\nsrc/\n├── commands/ # parses user actions\n├── sessions/ # owns session state\n└── transport/ # sends API requests\n```\n\n- Show component interaction, control flow, or data flow with Mermaid:\n\n```mermaid\nsequenceDiagram\n participant User\n participant UI\n participant Daemon\n User->>UI: choose command\n UI->>Daemon: send expanded prompt\n Daemon-->>UI: stream result\n```\n\n- Use `diff` when the point is what changes and the surrounding shape already exists. Match the diff shape to the topic.\n\nFor a component change:\n\n```diff\n \n useSessionEvents()\n \n+ \n \n+ \n```\n\nFor a file-layout change:\n\n```diff\n src/\n ├── commands/\n+│ └── show-me.ts # expands the slash command\n ├── sessions/\n-└── transport.ts\n+└── transport/\n+ ├── client.ts\n+ └── stream.ts\n```\n\nFor a call-tree or call-stack change:\n\n```diff\n submitForm\n createSession\n persistPrompt\n+ expandSkillMention\n launchAgent\n- navigateToSession\n+ navigateToSession\n+ subscribeToEvents\n```\n\nFor a state or control-flow change:\n\n```diff\n on(save)\n- write content\n+ if content is unchanged\n+ return cached result\n+ write new content\n+ invalidate cache\n```\n\n- Show the whole block when most of it is new, when omitted context would hide ownership or order, or when the user needs a copyable target shape:\n\n```ts\nfunction expandSkill(command: string): string {\n const skillName = command.slice(1);\n return `use the ${skillName} skill`;\n}\n```\n\n#### Guidance\n\nPlace each visual next to the short text it supports. Keep only the calls, files, props, states, and boundaries needed to answer the user's current question or the options to resolve the current discussion point.\n\nYou may use one of these, you may use several, it is unlikely you will use all of them. Use your judgement and don't overwhelm the user.\n\n### Evidence\n\nConcrete evidence that the change works. Show a before and after.\n\nScreenshots are S-tier - when the environment is set up for it and the change is visual.\n\nExecution-based evidence is A-tier. Test results, console output. Show the exact test that now fails and passes, using pseudocode.\n\n### Merge Danger\n\nDescribe whether it's a one-way or two-way door. You can walk back through two-way doors, but not one-way doors. A PR that is cheap to roll back is lower risk. Changes that involve destructive actions or hard-to-reverse decisions are one-way doors.\n\nThe blast radius is the potential impact or scope of the changes introduced by this PR. Consider all possibilities. Examples are layout shift, breakages for consumers, mobile responsiveness, etc.\n","frontmatter":{"name":"pr","description":"Use when writing a PR body.","metadata":{"credits":{"skill":"show-me","author":"Dex Horthy","organisation":"Humanlayer","url":"https://github.com/humanlayer/skills/blob/main/plugins/show-me/skills/show-me/SKILL.md"}}},"isInternal":false,"tokens":980,"sizeBytes":4169},{"name":"openai.yaml","path":".agents/skills/prototype/agents/openai.yaml","rawUrl":"https://raw.githubusercontent.com/withcoral/coral/HEAD/.agents/skills/prototype/agents/openai.yaml","title":"Agents Skill","category":"anthropic-skill","format":"yaml","content":"interface:\n display_name: \"Prototype\"\n short_description: \"Prototype to answer a design question\"\n","isInternal":false,"tokens":21,"sizeBytes":100},{"name":"LOGIC.md","path":".agents/skills/prototype/LOGIC.md","rawUrl":"https://raw.githubusercontent.com/withcoral/coral/HEAD/.agents/skills/prototype/LOGIC.md","title":"Prototype Skill","category":"anthropic-skill","format":"markdown","content":"# Logic Prototype\n\nA single, self-contained HTML file (a **shareable demo**) that lets anyone drive a state model by clicking buttons. Use this when the question is about **business logic, state transitions, or data shape**: the kind of thing that looks reasonable on paper but only feels wrong once you push it through real cases.\n\nBecause it's one file with nothing to install, you can hand it to a non-developer (a designer, a PM, a domain expert) and let them feel the model for themselves. So it speaks their language, not the code's.\n\n## When this is the right shape\n\n- \"I'm not sure if this state machine handles the edge case where X then Y.\"\n- \"Does this data model actually let me represent the case where...\"\n- \"I want to feel out what the API should look like before writing it.\"\n- Anything where someone wants to **press buttons and watch state change**.\n\nIf the question is \"what should this look like,\" this is the wrong branch. Use [UI.md](UI.md).\n\n## Process\n\n### 1. State the question\n\nBefore writing code, write down what state model and what question you're prototyping. One paragraph, at the top of the demo (in a visible intro, not just a comment). A logic prototype that answers the wrong question is pure waste, so make the question explicit so it can be checked later, whether the user is watching now or returning to it AFK.\n\n### 2. Isolate the logic in a portable module\n\nPut the actual logic (the bit that's answering the question) in a single `