codebase-memory-mcp Documentation

Technical Documentation: DeusData/codebase-memory-mcp

ℹ️ Provenance: Multi-Tier Ingestion · CodeWiki Reference

Overview (README.md) Copy Section

codebase-memory-mcp: local code intelligence for AI agents via MCP

Understand your repository, from source code to the systems around it. codebase-memory-mcp is an MCP server that connects code, documentation, configuration, and infrastructure in a persistent local knowledge graph. Use it to find an unfamiliar feature, follow a call across services, trace a deployment back to its source, or assess which parts of a project a change could affect.

Give your coding agent structural context through Model Context Protocol (MCP), or query the graph directly from the CLI. The MCP server is the core of the project. Its optional visual companion, CBM Lens, is in open beta (v0.0.1), with a source reader, architecture maps, the Galaxy graph, coverage diagnostics, architecture decisions, and local code chat.

The native C backend exposes 17 MCP tools, parses languages and file formats with vendored grammars, and adds Hybrid LSP type resolution for supported languages. The installer supports 45 automatic or conditional client integrations, including client configuration, skills, agent profiles, and applicable hooks. Indexing and graph queries need no hosted service or API key. Focused graph queries can reduce repeated file reads and token use; see the measured results.

Quick start · Capabilities · Documentation · Infrastructure · Smart CI · Explore CBM Lens · Agent setup · MCP tools

Local processing: Source analysis and graph storage run on your machine. CBM Lens can optionally download a model for browser AI after you enable it; inference stays on the device. The installer writes agent configuration for the clients you select. See Security for release verification and SECURITY.md for reporting issues.

See the graph with CBM Lens (open beta v0.0.1). Folders become nested platforms, files and symbols become selectable blocks, and arrows show indexed relationships. The recording opens src/foundation in the map of this repository, selects arena.c and opens its source. Height can show source size; hotspot gravity reflects static fan-in. Still image.

Galaxy. Explore the code knowledge graph in 3D. The recording follows the outgoing relationships of cbm_arena_alloc one layer further, switches to the hierarchy and shows the two-hop call path to cbm_alloc. Edge pulses show relationship direction; they do not represent live execution. Still: two dependency layers around cbm_max_file_bytes.

Explore. Browse files in an IDE-style workspace, read syntax-highlighted source in Monaco, and keep the code graph alongside it. File impact shows what depends on the open file, and Code details shows the callers and calls of the symbol under the caret. Select code to ask the local chat about it.

More views: system structure and behavior

System structure groups symbols into components and lets you inspect their dependencies.

Behavior shows possible calls from a selected operation and the source evidence behind them.

The high-resolution GIFs show CBM Lens Beta on 20 October 2026; the still images were captured on 12–13 October. All show public source from this repository. See capture details for each capture's build, indexed scope and recording setup.

Why codebase-memory-mcp Copy Chapter

• Find the right code with context: combine symbol, text, and semantic search with callers, dependencies, source snippets, and file outlines. Targeted queries and compact, paged results can reduce token use and repeated file reads; see Performance for scoped measurements.
• Connect documentation to implementation: index document sections, PDF text pages, architecture decisions, and supported references into code. Find the explanation alongside the symbol.
• Follow the system beyond source files: trace supported relationships through Docker, Kubernetes, Helm, Kustomize, Terraform/OpenTofu, Terragrunt, Ansible, scripts, Make, and CI configuration.
• Understand a change before running it: inspect affected symbols, compare indexed snapshots, and find related tests. Experimental Smart CI combines static relationships and per-test coverage for supported C/C++ test harnesses, falling back to broader runs when evidence is insufficient.
• Explore visually with CBM Lens Beta: read source, follow call paths, inspect architecture and coverage gaps, manage architecture decisions, and ask a local model about selected code.
• Equip your coding agents: install sets up supported client configuration, durable guidance, skills, Scout/Verify/Auditor profiles, and applicable context hooks. Support varies by client; see the integration matrix.
• Keep context local and reusable: indexes persist between sessions, eligible active MCP sessions can keep them synchronized, and optional graph artifacts let a team share a baseline.
• Run a native tool on your machine: standard releases include CBM Lens for macOS, Linux, and Windows. The native runtime needs no Docker or language runtime; parser grammars are included.

Quick Start Copy Chapter

Install the latest release (macOS / Linux):

curl -fsSL https://raw.githubusercontent.com/DeusData/codebase-memory-mcp/main/install.sh | bash

The standard native release includes CBM Lens. Start it after installation with codebase-memory-mcp daemon start --open --port=9749. The Nix packages and source builds also offer builds without the UI.

Windows (PowerShell):

# 1. Download the installer
Invoke-WebRequest -Uri https://raw.githubusercontent.com/DeusData/codebase-memory-mcp/main/install.ps1 -OutFile install.ps1

# 2. (Optional but recommended) Inspect the script
notepad install.ps1

# 3. Unblock the downloaded file (removes Mark-of-the-Web restriction added by browsers/Invoke-WebRequest)
Unblock-File .\install.ps1

# 4. Run it
.\install.ps1

Note: If you see a script execution policy error, run Set-ExecutionPolicy -Scope Process Bypass first, or invoke with PowerShell -ExecutionPolicy Bypass -File .\install.ps1.

Options: --skip-config (binary only, no agent setup), --dir= (custom location).

Antivirus note: Microsoft Defender may flag a release binary as
Trojan:Script/Wacatac.B!ml. This is a known false positive; typically 61 of
~62 engines return clean, and the same detection family hits gh, llama.cpp,
Godot and Microsoft's own Go toolchain. See
Antivirus False Positives for the
evidence, how to verify the artifacts yourself, and how to report it if you
think we are wrong.

Restart your coding agent. Say "Index this project": done.

Manual install

1. Download the archive for your platform from the latest release:

   • codebase-memory-mcp-<os>-<arch>.tar.gz (macOS/Linux) or .zip (Windows)

2. Extract and install (each archive includes install.sh or install.ps1):

   macOS / Linux:

   bash Copy

   tar xzf codebase-memory-mcp-*.tar.gz
   ./install.sh

   Windows (PowerShell):

   powershell Copy

   Expand-Archive codebase-memory-mcp-windows-amd64.zip -DestinationPath .
   Unblock-File .\install.ps1
   .\install.ps1

3. Restart your coding agent.

The install command automatically strips macOS quarantine attributes and ad-hoc signs the binary; no manual xattr/codesign needed.

Agent setup: skills, profiles, and hooks Copy Section

Installation also equips supported coding agents to use the graph in daily work. The native install command detects clients and installs the integration pieces their documented formats support:

Hook augmentation is context-only and fail-open: it does not replace or block the requested tool call. Features vary by client and platform; some clients use durable instructions or a parent-agent handoff instead of hooks or direct subagent MCP access. Codex users review and trust installed hooks through /hooks. See Multi-Agent Support and Sessions, compaction, and subagents for exact support and activation requirements.

Session Coordination Daemon Copy Section

CBM automatically shares one per-account coordination daemon across Claude Code, Codex, OpenCode, and every other configured client. There is no opt-in setting for MCP servers or hook clients: the first daemon-backed CBM session starts it, each session registers its own work, and a session-managed daemon stops after the final session. A daemon started with daemon start remains running until stopped. The daemon owns long-lived background services such as watchers, shared indexing jobs, and the optional UI. Closing one session cancels work owned only by that session, while work still needed by another session continues.

The detached daemon does not depend on an MCP frontend's stderr. It keeps owner-only durable records under the canonical ${CBM_CACHE_DIR}/logs directory (default ~/.cache/codebase-memory-mcp/logs):

Thin frontends still write immediate startup and session-specific errors to their own stderr; MCP JSON-RPC stdout remains clean.

All active CBM processes must run the exact same version, executable build, coordination ABI, and canonical cache root. Equivalent CBM_CACHE_DIR aliases resolve to the same root; a genuinely different root is rejected while any CBM process is active. MCP servers, hooks, one-shot CLI commands, temporary index workers, and the daemon share a crash-safe OS admission barrier; starting an ordinary conflicting process fails before doing work and records an explicit conflict in ${CBM_CACHE_DIR}/logs/daemon-conflicts.ndjson.

The native install and uninstall commands are the deliberate exception to that conflict rule. For installation, download, verification, and private same-filesystem staging finish before activation so a bad candidate never disrupts active work. The native update command prints the installer command that performs the update. Activation then publishes account-wide maintenance intent, asks the daemon and every temporary local operation to cancel, and waits to a finite deadline for all coordinated CBM processes to exit. It holds the admission and lifetime barriers exclusively while changing the active binary, configuration, PATH, or indexes. New CBM work cannot enter during this window. Activation progress and results are recorded in ${CBM_CACHE_DIR}/logs/activation-events.ndjson, and a successful command tells you to restart open coding-agent sessions so they launch the activated build.

Package-manager setup (npm, PyPI, or Go) verifies and caches the native executable for that package. It does not replace the active native installation and therefore does not stop running CBM sessions. When that cached binary is executed, it still enters the same exact-build admission barrier. The shell and PowerShell installers invoke the verified candidate's native install command, so they do receive the full account-wide activation guarantee.

The ordinary cli mode runs one command locally, but it is not disconnected from the daemon above: it connects to the shared coordination daemon (starting one if none is already running) for that same admission barrier plus per-project locks on graph mutations. Its connection is a cli_session: it does not register a project watcher or open a browser, and it holds the admission lease only for the command's lifetime. The shared daemon may serve Lens according to its UI configuration and available assets. If the command's own connection is what started the daemon, that daemon exits again once the command closes and no other session is attached; if a daemon was already running, the command simply joins and leaves it exactly as found. While the command is running, a temporary monitor separately lets activation cancel that operation and its supervised worker safely; the monitor exits with the command and never becomes a standing daemon. See CLI Mode for details.

CBM Lens Beta: code visualization and source navigation Copy Section

CBM Lens is an optional visual companion to the MCP server, currently in open beta (v0.0.1). Standard native releases include its browser interface, so you can launch Lens directly after installation; UI-enabled source builds include it too. It builds on CodeAtlasWeb by Bernhard Jackiewicz (design and tracking in #1964). Lens is served on loopback. Browser AI downloads its optional model only after you enable it; inference stays on-device. Each panel names where its facts come from, and what the index did not record is shown as a gap. Start the local server:

codebase-memory-mcp daemon start --open --port=9749

The browser opens at http://127.0.0.1:9749. A permanent daemon keeps the UI available after the command exits. Use daemon status to inspect it and daemon stop when you are finished. For an MCP session, --ui=true --port=9749 enables the same shared UI. Active clients must use the same executable build and cache directory.

Open your first repository

1. Open the project menu and choose Add project index. You can also use Alt+P.
2. Choose a local repository folder and name the index. Files stay in place.
3. Start indexing and follow its reported status. You can close the dialog while it runs and reopen its activity from the project menu.
4. Open the indexed project. Start with Architecture for an overview or Explore to read a file.

To bookmark a view, use ?project=<name>&workspace=architecture. Other workspace values are explore, galaxy, adr, coverage, and system.

Architecture views

• Overview: move through source folders and files on a 3D map. Choose source-size or uniform height, language or kind colors, and relationship filters. Switch to Entry points to follow a static call graph from a starting symbol.
• Routes: switch between the service map and endpoint connections. Deployment metadata and indexed route evidence are shown separately; unresolved registrations remain visible.
• Hotspots: inspect symbols using measured signals such as fan-in and complexity. Gravity wells help locate frequently referenced areas. They are not a code-quality score.
• System structure: inspect component dependencies, expand groups, isolate cycles, and follow connections back to source evidence.
• Behavior: choose a source and a reachable destination to inspect static call paths and their source statements. These paths show possible relationships, not a recorded execution order. Runtime observations require separately ingested traces.

Drag a 3D view to orbit, scroll to zoom, and select a node or edge to inspect it. Edge colors distinguish relationships, and pulses indicate their direction. Reduced-motion preferences pause the pulses. Large views show their limits and omitted counts. A missing edge can reflect an indexing or display limit, so check Coverage and the source before drawing conclusions.

Chat with your codebase locally

Open Chat with the panel button at the top right to ask about the code you are looking at. In Explore it uses the open file or the exact text you have selected; in Galaxy it uses the selected node and its loaded neighbors. You can also attach graph evidence through Add context.

1. Click the robot icon at the top right to open the agent configuration.
2. Choose a local model. The default Qwen2.5 Coder 0.5B downloads about 567 MB; Qwen3 0.6B, LFM2.5 1.2B and Qwen3.5 2B are larger options. All need WebGPU with shader-f16 support.
3. Click Download & load. Model files come from Hugging Face; inference runs in your browser, with no hosted chat API or API key. A model that is already in the browser cache loads with Load model (cached, no download).
4. Ask a question. The composer shows the file or selection that will accompany it. Select a smaller range when you want to focus the answer.

Questions about callers and callees, such as "Who calls cbm_pipeline_run?", are answered from the loaded graph without the model, and the answer says so. When you select a symbol, the chat lists its indexed facts and the lines of its code, and the model adds one sentence; that sentence is left out when it claims something the source does not show. Text written by the model is labelled as such.

The recording asks who calls arena_grow and what it calls, with both answers listed directly from the indexed graph beside the selected symbol. Still image of a model explanation and graph answer.

The chat history is kept for 24 hours in this browser, per project. Each new message uses the source currently open in the reader; earlier source snapshots are not silently accumulated, and messages about another file or selection are not sent along. If a prompt is too long for the model, older messages are left out first; if it still does not fit, CBM Lens asks you to select less code instead of cutting the code. This chat explains the context you give it; it cannot edit files or run tools. Check its answers against the source. The small default model words things loosely, especially in languages other than English.

Search

Cmd+K or Ctrl+K opens the Galaxy search. It finds symbols, files and folders by name or path in the whole index, also nodes that are not loaded yet. For callers and callees, ask in the chat; the answer is listed from the graph.

Architecture decisions

The ADR workspace keeps an unsaved draft per project in this browser tab. It checks whether the stored document changed before saving, but the API does not provide an atomic document lock. The editor accepts up to 8,000 UTF-8 bytes. Saving changes the indexed decision document, not a file in your repository.

Experimental agent activity

The Agents workspace and activity overlays are hidden in normal builds. To include them in a source build:

VITE_EXPERIMENTAL_AGENTS=true scripts/build.sh --with-ui

Activity comes from explicitly configured tool hooks. The bundled adapter supports Claude Code PostToolUse; other clients need their own adapter. Recorded events go to the local daemon, and the reader must enable activity before the browser polls them. This does not automatically observe every agent. See the frontend setup guide.

For troubleshooting, open System logs. Frontend errors and failed requests are recorded locally. GET /api/ui-log reads frontend entries from the daemon's SQLite journal; ${CBM_CACHE_DIR}/logs/ui.log is also retained as a rotating JSONL export. See the frontend README for log scope, model setup, and development details.

Auto-Index Copy Section

Enable automatic indexing on MCP session start:

codebase-memory-mcp config set auto_index true

auto_index is off by default. When enabled, new projects are indexed automatically on first connection. The configurable file limit is config set auto_index_limit 50000. Previously indexed projects can be watched independently through auto_watch (on by default), subject to the session rules below.

The watcher follows MCP sessions: it watches the project each open MCP session is rooted in (the client's working directory) and stops when the last session for that project closes. A repository indexed with cli index_repository, or indexed from a session rooted somewhere else, is not watched; re-run index_repository after changing it. index_status reports this in its watch object: watched, and either the poll cadence and last scan time or the reason it is not watched.

Watcher registration is controlled separately by auto_watch (default true). Set config set auto_watch false to keep a session from registering its project with the background watcher; useful when working across many projects and you want each session contained to explicit indexing.

To turn the watcher off entirely, set config set watcher_enabled false (default true): the background poll thread never starts and no project is registered, while auto_index and manual index_repository keep working. Unlike auto_watch, which is consulted per session, watcher_enabled is read once when the background daemon starts, so run codebase-memory-mcp daemon stop after changing it; reconnecting your MCP client alone will not restart the daemon. See docs/CONFIGURATION.md.

The watcher follows git projects only by default: a project indexed from a directory that is not a git repository is not watched and stays at its last index until you run index_repository again. To have such roots polled too, set config set watch_non_git true (default false, read when the daemon starts): they are then checked on the same cadence with a file-tree scan that uses the indexer's own ignore rules.

Keeping Up to Date Copy Section

Updates run from the install script on every platform, not from inside the running binary. codebase-memory-mcp update validates your flags and then prints the exact command to run:

# macOS / Linux
bash "<install-dir>/install.sh"

# Windows
powershell -File "<install-dir>\install.ps1"

The install script is placed next to the binary at install time, so the printed path resolves beside the executable. It is idempotent, so re-running it is the update: it stops the daemon, retires the running binary, installs the new one, and cleans up.

Why it works this way. On Windows it is a hard requirement; a running executable cannot replace its own image, so the swap has to happen from a process that is not the binary being replaced. On macOS and Linux it is a deliberate choice: an in-process updater is structurally a downloader (fetch an archive, verify it, unpack it, mark a file executable, run it), and shipping that composite in every binary to serve a command most people run a handful of times is a poor trade. The release archives now carry no download URLs at all, and cbm makes no network request of its own accord: it does not check for new versions in the background, and nothing phones home. You find out about releases from the install script, your package manager, or GitHub.

If PowerShell refuses to run the script because the file came from the internet, Unblock-File it first.

Installed through npm or pip? Update with your package manager on every platform (npm install codebase-memory-mcp@latest / pip install -U codebase-memory-mcp).

Uninstall Copy Section

codebase-memory-mcp uninstall

Removes owned agent config entries, skills, hooks, instructions, and the installed binary. Existing graph indexes are listed and kept by default. -y/--yes, -n/--no, and noninteractive input keep them unless --delete-indexes is explicitly given. An interactive terminal without that flag is asked separately; the default answer is to keep them.

codebase-memory-mcp uninstall -y --delete-indexes   # also delete every project index
codebase-memory-mcp uninstall --dry-run --delete-indexes   # preview; change nothing

--delete-indexes is explicit consent and takes precedence over --no; --dry-run always preserves the files. When indexes are kept, uninstall prints their cache directory (${CBM_CACHE_DIR:-~/.cache/codebase-memory-mcp}) and how to remove them.

The install script placed beside the binary is reported, not deleted: uninstall prints its path and the rm command for it. It is left alone on purpose: it may be your own copy, a symlink into a checkout, or managed by a package manager, and an uninstaller should not delete a file it cannot prove it owns.

Features Copy Chapter

Graph & analysis Copy Section

• Architecture overview: get_architecture returns a compact summary by default. Request specific aspects for routes, hotspots, dependencies, clusters, or cycles. system_structure provides the projection used by CBM Lens and must be requested on its own.
• Architecture Decision Records: manage_adr persists the project's decision record across sessions; indexing also extracts supported ADR metadata and supersession from documentation files
• Louvain community detection: Discovers functional modules by clustering call edges
• Change impact: detect_changes maps uncommitted changes to affected symbols with risk classification and can report test impact with scope: "tests"; compare_graphs compares indexed snapshots
• Runtime observations: ingest_traces adds observed call relationships from separately collected traces alongside static analysis; static call paths alone do not show what ran
• Call graph: Resolves function calls across files and packages (import-aware, type-inferred)
• Dead code candidates: find functions with no indexed callers, excluding entry points. Check dynamic dispatch, coverage gaps, and framework conventions before removing code.
• Cypher-like queries: MATCH (f:Function)-[:CALLS]->(g) WHERE f.name = 'main' RETURN g.name

Search Copy Section

• Semantic search (the semantic_query parameter of search_graph, not a separate tool): vector search across the entire graph, powered by bundled Nomic nomic-embed-code embeddings (40K tokens, 768d int8) compiled into the binary; no API key, no Ollama, no Docker. 11-signal combined scoring (TF-IDF, RRI, API/Type/Decorator signatures, AST profiles, data flow, Halstead-lite, MinHash, module proximity, graph diffusion).
• BM25 full-text search via SQLite FTS5 with cbm_camel_split tokenizer (camelCase / snake_case aware)
• Structural search (search_graph): regex name patterns, label filters, min/max degree, file scoping
• Code search (search_code): graph-augmented grep over indexed files only

Documentation connected to code Copy Section

Find the explanation behind a symbol and follow supported references back to its implementation. Documentation participates in the graph alongside source files:

• Markdown, reStructuredText, and AsciiDoc become navigable document sections; PDFs with text layers become searchable text pages. PDF extraction does not perform OCR.
• Document-to-code references and supported language doc-comment references create MENTIONS relationships when the target can be resolved.
• Architecture decision records contribute metadata such as status, deciders, and supersession, so decisions can be explored with the code they describe.
• Supported diagrams and OpenAPI descriptions contribute structured entities. Extraction and reference resolution differ by format; not every textual mention becomes a graph edge.

Use graph and text search to locate the relevant section, follow its recorded relationships, and open the source evidence. Check check_index_coverage before treating a missing result as proof that the repository has no such documentation or reference.

Infrastructure and deployment Copy Section

Ask "What deploys this service?", "Which image runs it?", or "Where does this configuration reach code?" CBM connects supported infrastructure forms to the source they build, configure, and run:

• Containers and clusters: Dockerfiles, Compose services, Kubernetes resources, Helm charts and releases, and Kustomize overlays and patches.
• Provisioning and configuration: Terraform/OpenTofu resources and modules, tfvars overrides, Terragrunt dependencies, Ansible tasks and handlers, and supported configuration keys and environment-variable access.
• Build and delivery: shell commands, Make targets, and supported CI steps that build images, invoke repository scripts, or deploy resources.

For resolvable inputs, a trace can follow Service → workload → image → Dockerfile → program, or follow deployment and configuration relationships. Use trace_path modes for topology, deployment, and run paths. These are static relationships from supported source forms; dynamic values and unresolved templates can leave gaps. See Edge types for the exact relationship rules and limitations.

Cross-service linking Copy Section

• HTTP route ↔ call-site matching with confidence scoring
• gRPC, GraphQL, tRPC service detection with protobuf Route extraction
• Channel detection (EMITS / LISTENS_ON) for Socket.IO, EventEmitter, and generic pub-sub patterns across 8 languages with constant resolution

Cross-repo intelligence Copy Section

• CROSS_* edges link nodes across multiple repos indexed under the same store
• Cross-repo architecture summary combining services, routes, and dependencies across the indexed fleet
• Run status: index_status (and the UI layout next to linked_projects) carries a cross_repo object, {"status":"never_run"} until index_repository(mode="cross-repo-intelligence") has run with that project as the source, then "ran" with last_run_at, outcome, targets, projects_scanned, total_cross_edges and skipped_projects. That is how you tell "never linked" apart from "linked, nothing matched". The status is stored in the project's own index next to its CROSS_* edges, so a reindex that rebuilds the index drops both at once and the status goes back to never_run.
• Skipped stores: target_projects: ["*"] skips an index built before the #768 schema change rather than failing the whole run. The response lists each one as skipped_projects: [{"name", "reason": "pre_768_schema", "hint": "reindex this project"}].

Smart CI and test impact Copy Section

Use the graph to explain which tests a change reaches, and use that evidence to focus supported CI runs. Smart CI is experimental and has two layers:

codebase-memory-mcp test-impact select produces a selection with reasons; test-impact publish produces the graph-and-coverage bundle used as its baseline. For agent-assisted investigation, detect_changes with scope: "tests" exposes test impact through MCP.

Narrowing currently supports C/C++ family code with this repository's test conventions or declared equivalents. It is not general test selection for every language CBM indexes. Missing, stale, incompatible, or incomplete evidence broadens the selection or returns run_all; static reachability without admitted coverage never omits a suite. Begin with report-only shadow mode and validate your harness before enabling gating.

In this repository, test narrowing is wired to the Linux x86_64 test leg when TEST_IMPACT_MODE=gate; release and dry-run workflows retain full execution. See the Smart CI guide for supported conventions, coverage admission, setup, version requirements, and fallback behavior.

Edge types (selected) Copy Section

• CALLS: a callable is invoked at the source site
• CALL_REFERENCE: a callable is used at a supported reference site (for example, a direct value argument) and resolves to one exact target
• USAGE: an identifier is used, but a unique callable target is not proven (including ambiguous or complex expressions)
• IMPORTS, DEFINES, IMPLEMENTS, INHERITS
• HTTP_CALLS, ASYNC_CALLS (cross-service); a shell script, Make recipe or CI step running curl / wget on a URL it states (constant words resolved, unknown parts {}) to its Route with the method its flags give: a Kubernetes Service of the repository by the path, as code clients do, any other host (a download, localhost) with the host kept; the edge keeps via, method, url_path, host, never the user info or the query
• SPAWNS: the caller starts another program (subprocess.run, exec.Command, posix_spawn, Process.Start, a shell command or Make recipe running a program by path, ...); the target is a Process node named after the program, or <dynamic> when the program is not a literal. When a shell script, Make recipe, GitHub Actions step or Prow job runs a script of the repository (./hack/x.sh, bash hack/x.sh), the target is that script's Module (else its File); make TARGET (also $(MAKE), -C, -f) targets the rule of that target, followed through included makefiles; edges from CI carry ci and job. A Dockerfile's last stage ENTRYPOINT (else CMD) spawns what the container runs (api: dockerfile): the repository script or source copied into the image, the Go main of a binary a stage builds with go build -o, else a Process
• INFRA_MAPS: a Kubernetes Service to each workload whose pod labels carry every key: value of the Service's selector, within one namespace
• CONNECTS_TO: a Kubernetes workload or compose service to the Service its containers' environment addresses (PRODUCT_CATALOG_SERVICE_ADDR: productcatalogservice:3550, a URL's host, <svc>.<ns>.svc..., a bare name on an *_ADDR / *_HOST / *_URL / ... key; compose: the project's .env interpolated): a Kubernetes Service of the same bundle and namespace, or a compose service of the same file. The edge keeps via: env, key and container, never the value
• USAGE with ref: image: a Kubernetes workload (any pod template: containers, initContainers, CronJobs, custom resources) or a docker compose service to the Image it runs
• Helm templates are rendered for their images: with the chart's values (the Chart DEPLOYS them, default_of: the values keys read) and with each release's (a Flux HelmRelease, a kustomize helmCharts entry, a Terraform helm_release: override_of). An image the render cannot know is left out, never guessed: after a change of values it does not run exactly, a tpl or include it cannot look into, a release value it cannot read, or a work budget it runs past. Known limit: each template file renders on its own, while Helm renders a chart's files with one shared .Values (a subchart's is its parent's .Values.<alias>), so a chart that changes .Values in one file and reads it in another can still show a wrong image
• DEPLOYS: what puts manifests, kustomizations or charts on a cluster, to what it puts there: a Flux Kustomization or Argo Application to the kustomizations / manifests / chart of its path, a Flux HelmRelease to its chart (spec.chart or chartRef → OCIRepository), a shell command or Make recipe (kubectl apply -f/-k, kustomize build | kubectl apply, helm install/upgrade/template), a Go installer loading a chart with the Helm SDK, a Terraform helm_release, a Terraform kubernetes_* resource the object it declares (a Resource like a manifest's, ecosystem: terraform, when its name is a literal), a kustomization inflating a chart (helmCharts; its values file an IMPORTS). A chart outside the repository is an external Chart, keyed by its publisher when its repository URL is stated (__chart__<repository>/<name>, the same node in every project, the URL in repository), else one per project and name; the edge says via (flux, argo, helmrelease, kubectl, kustomize, helm, helm-sdk, terraform, helmcharts, ...)
• DEPENDS_ON: a Helm chart to the charts it declares (the local chart for a resolvable file:// repository, otherwise an external Chart; the edge keeps repository), and a go.mod / requirements.txt file to an external Package; a Dockerfile to the Images it builds on (via: from, copy-from); an Image to the Dockerfile it is built from (via: build; tool: skaffold, compose, docker-build for docker / podman / nerdctl / buildah build commands in scripts, Make and CI); a compose service to the services its depends_on / links name
• CONFIGURES: code reading an environment variable to the shared EnvVar of that name (strategy: env_access); a Kubernetes container, compose service or Dockerfile (ENV of the image) setting one to that container's own EnvVar (strategy: env_set, with source: value, configMapKeyRef, secretKeyRef, fieldRef, ..., compose, dockerfile; values are never stored); and that container's EnvVar to the code reading the name inside the files its image is built from (strategy: env_read, with dockerfile): the image's files are the sources its Dockerfile's COPY / ADD put into it, less the context's .dockerignore
• CONFIGURES with strategy: key_symbol: a code symbol to the config key it names (config_key). A YAML, JSON, HCL (Terraform, OpenTofu, tfvars), TOML or INI key links by exact name (maxRetryCount = max_retry_count) and only to code inside the config file's component: its directory, or the one around a config/, configs/, conf/, settings/ or etc/ directory. A .env or .properties key also links by substring (confidence 0.75)
• OVERRIDE with via: patch: a Kustomize patch to each object of its kustomization's build it changes (child → base): a patch file's documents to the object of the same kind and name; an inline patch (from the kustomization) or a JSON6902 operation file to what its target selects (group, version, kind, name, namespace, label selector); a component's in the build of each kustomization including it. The edge keeps kustomization, patch and line; a patch file is an IMPORTS (local_name: patches) of its kustomization, never a member of the build
• OVERRIDE with via: namePrefix / nameSuffix: an object a kustomization renames, from its variant: a Resource named with the deployed name (capi- + controller-manager; the prefixes of nested kustomizations outermost first, suffixes innermost first) in the namespace that build gives it, ecosystem: kustomize, which the kustomization's file DEFINES. CRDs, Namespaces and API services keep their names, as kustomize does. The edge keeps kustomization and line
• USAGE with via: terraform: a Terraform / OpenTofu reference (var.X, local.X, module.X, data.T.N, T.N) from the top-level block holding it to the block of its module (the .tf / .tofu files of its directory) that declares it (a locals entry: its locals block); an entry of the block's depends_on list is a DEPENDS_ON instead. The edge keeps ref (var, local, module, data, resource) and line; a name only mentioned in string text is no reference. A module call with a local source (module "net" { source = "./modules/net" }) IMPORTS (local_name: the module's name) the Folder of that directory (the root directory: the Project), and module.net.id (also module.net[0].id, module.net[*].id) USAGEs (ref: output) the output "id" there, beside module.net
• OVERRIDE with via: tfvars: a .tfvars assignment to a variable the module of its own directory declares, from the file's Variable to that variable block (the edge keeps line); a .tfvars file whose own directory declares none of its names sets the nearest ancestor module's that does, the edge marked inferred (such a file is applied by a -var-file the repository does not state)
• Terragrunt (via: terragrunt): a unit (terragrunt.hcl), evaluated in its own context: an include IMPORTS (local_name) the file its path names; a read_terragrunt_config IMPORTS (read: the file holding the call) the file it names, from the block holding the call or, for a call in an included file, from the unit's include block that brings it in; a dependency DEPENDS_ON the other unit; a local terraform source IMPORTS its Folder. Paths may use find_in_parent_folders, dirname and get_terragrunt_dir; anything else is unknown and links nothing
• Ansible (via: ansible): a playbook or task file (a YAML file whose top level is a sequence) IMPORTS (local_name: the directive) what its directives name: import_playbook a playbook, include_tasks / import_tasks / include a task file (from its directory, else its role's tasks/), roles: items and include_role / import_role a role's Folder (roles/<name> at or above its directory). A templated target links nothing. Each task and handler is a Function (named by its name:, else its module; source: ansible, kind, module) that its file DEFINES: in a playbook's plays, a role's tasks/ and handlers/ files, and the files those include; a block's tasks, not the block. A notify CALLS (via: notify) each handler of that name or listen topic in the task's own role (else its file). A kube (kubespray's module) or kubernetes.core.k8s task DEPLOYS (via: ansible, module) the repository file its play rendered the applied manifest from: the template (in the role's templates/) or copy (its files/) task whose dest is the task's filename, item by item through register / with_items / loop lists the play states (a literal list, a set_fact, the role's defaults / vars), or a k8s template: argument
• EMITS, LISTENS_ON (channels)
• DATA_FLOWS with arg-to-param mapping + field access chains
• SIMILAR_TO (MinHash + LSH near-clone detection, Jaccard scored)
• SEMANTICALLY_RELATED (vocabulary-mismatch, same-language, score ≥ 0.80)

Indexing pipeline Copy Section

• 162 vendored tree-sitter grammars compiled into the binary
• Generic package / module resolution: bare specifiers like @myorg/pkg, github.com/foo/bar, use my_crate::foo resolved via manifest scanning (package.json, go.mod, Cargo.toml, pyproject.toml, composer.json, pubspec.yaml, pom.xml, build.gradle, mix.exs, *.gemspec)
• Infrastructure-as-code indexing: Dockerfiles (Dockerfile.*, *.Dockerfile, Containerfile: bases, what the container runs, ENV), container images (Image: the normalised repository, tag and digest on the edges; a repository built from several Dockerfiles, one per component tag, gets an Image per component), docker compose services (Resource), Kubernetes manifests (Resource, with group, version, kind, namespace, labels as k=v,k2=v2), Kustomize overlays (IMPORTS to their entries and patch files, OVERRIDE from their patches), Helm charts (Chart, DEPENDS_ON), Terraform / OpenTofu (.tf, .tofu, .tfvars) as graph nodes. A Service walks to the code it runs: Service → workload → Image → Dockerfile → main
• Hybrid LSP semantic type resolution for Python, TypeScript / JavaScript / JSX / TSX, PHP, C#, Go, C, C++, Java, Kotlin, Rust, and Perl; a lightweight C implementation of language type-resolution algorithms, structurally inspired by and compatible with major language servers including tsserver / typescript-go, pyright, gopls, Roslyn, Eclipse JDT, and rust-analyzer (parameter binding, return-type inference, generic substitution, JSX component dispatch, JSDoc inference for plain JS files, namespace + trait + late-static-binding resolution for PHP, file-scoped namespaces + records + LINQ method syntax for C#, class-hierarchy + overload + lambda resolution for Java, extension-function + scope-function resolution for Kotlin, trait-method + UFCS resolution for Rust)
• RAM-first pipeline: LZ4 compression, in-memory SQLite, single dump at end. Memory released after.

Distribution & operation Copy Section

• Native runtime set, zero infrastructure services: SQLite-backed, persists to ~/.cache/codebase-memory-mcp/
• Auto-sync: Background watcher detects file changes in the project an MCP session is working in and re-indexes it automatically (other indexed repos: re-run index_repository)
• Route nodes: REST endpoints are first-class graph entities
• CLI mode: codebase-memory-mcp cli search_graph --project my-project --name-pattern ".*Handler.*"
• Available on: npm, PyPI, Homebrew, Scoop, AUR, go install

Indexing modes Copy Section

Team-Shared Graph Artifact Copy Chapter

Share a compressed graph snapshot so teammates can reuse an indexed baseline.

.codebase-memory/graph.db.zst is a zstd-compressed snapshot of the knowledge graph that lives next to your source. The artifact is opt-in: index_repository writes it only with persistence: true (the default is false), and once it exists later indexes refresh it. Local indexes persist regardless. When a teammate indexes a clone with an existing artifact, CBM imports that baseline and applies their local changes incrementally.

• Format: SQLite database, indexes stripped, VACUUM INTO compacted, then zstd 1.5.7 compressed (8 to 13:1 ratio typical)
• Two tiers:
  • Best (zstd -9 + index strip + VACUUM INTO); written by index_repository with persistence: true
  • Fast (zstd -3); refreshes an existing artifact on every other index, including the watcher's low-latency incremental updates
• Bootstrap: when no local DB exists but the artifact is present, index_repository imports the artifact first, then runs incremental indexing, reusing the shared baseline where possible
• Merge policy: the first export creates .codebase-memory/.gitattributes with merge=ours. Check your repository's Git merge-driver setup before relying on it for concurrent artifact updates.
• Commit it deliberately: the artifact is rewritten on every index, including the watcher's Fast tier, and git stores each rewrite as a full new blob. Committing every refresh is what turns a 20 MB file into gigabytes of history; one team reached ~6 GB across ~350 commits of this single path. Pick a cadence (a release, a milestone, a nightly job) rather than committing every save.
• Git LFS, if it must move on every commit: track it from the repo-root .gitattributes and leave the auto-created .codebase-memory/.gitattributes in place; the nearer file goes on supplying merge=ours, and only filter comes from the root:
  gitattributes Copy

  .codebase-memory/graph.db.zst filter=lfs diff=lfs merge=lfs -text

  Track only the .zst; artifact.json is small and carries the schema version. The attribute applies to future commits only. Removing existing blobs requires a separate history migration. Two costs to weigh before adopting it: GitHub meters LFS storage and bandwidth, and its objects cannot be pruned without contacting support; and every teammate needs git lfs install; without it their checkout leaves a pointer file where the artifact should be, the integrity-checked import refuses it, and they fall back to a full reindex.
• Optional: shared export is opt-in, and Git commits remain your choice. Add .codebase-memory/ to .gitignore if you prefer everyone to reindex from scratch.

The artifact provides a shared baseline while each developer keeps a local query database.

How It Works Copy Chapter

codebase-memory-mcp is a structural analysis backend: it builds and queries the knowledge graph. The backend does not require a generative model. It relies on your MCP client (Claude Code, or any MCP-compatible agent) to be the intelligence layer.

You: "what calls ProcessOrder?"

Agent calls: trace_path(function_name="ProcessOrder", direction="inbound")

codebase-memory-mcp: executes graph query, returns structured results

Agent: presents the call chain in plain English

With MCP, your coding agent translates a question into tool calls. CBM Lens can also use an optional browser model for explanations; this does not change how the backend indexes or queries code.

Performance Copy Chapter

Published backend benchmarks on Apple M3 Pro. These are separate from the CBM Lens checks:

Memory-aware pipeline: Indexing keeps hot data in memory and can spill extraction results to disk when needed. Memory is released back to the OS after indexing completes.

More focused context, fewer tokens: The project's 31-repository preprint reports 10× lower token use and 2.1× fewer tool calls compared with file-by-file exploration. These March 2026 backend measurements cover the study's workloads and agent setup; the paper documents the full evaluation. An earlier five-query comparison reported ~3,400 versus ~412,000 tokens for structural queries. The study's system-performance measurements include the Linux indexing and Cypher timings above.

Explore the measurement guide to evaluate quality, latency and token use on your own workload.

Troubleshooting & Diagnostics Copy Chapter

codebase-memory-mcp indexes and queries locally and collects no telemetry. When you hit something we can't reproduce on our side (a slow memory climb over hours, a performance regression, a leak that only appears after days of real use), we have no diagnostic data unless you choose to send it. Here is how to capture it yourself.

Capture a diagnostics log Copy Section

Set CBM_DIAGNOSTICS=1 before the first daemon-backed MCP session starts, then reproduce the problem (let it run as long as it takes; a slow leak needs time to show in the trend). The shared daemon captures this setting from the session that starts it. If it is already running, close all daemon-backed sessions so it exits before changing the setting. The daemon creates a fresh owner-private cbm-diagnostics--<random> directory below the system temp directory ($TMPDIR or /tmp on macOS/Linux, %TEMP% on Windows). The exact paths are recorded by the diagnostics.start event in ${CBM_CACHE_DIR}/logs/cbm-daemon.log:

The private randomized directory prevents another local account from pre-placing a link or special file at a predictable diagnostics path. Its `` component is the shared daemon's process ID, also recorded by the daemon.start event. Set the variable consistently in the env block of each agent's MCP server config, or export it before launching the first session.

What to share Copy Section

When you open a memory/performance issue, attach the .ndjson trajectory: it contains no source code or query text, only resource counters. If you'd rather not attach a file, paste it (or an agent's summary of it) into the issue: your assistant can read the NDJSON directly and report whether rss/committed grow monotonically, how fast, and relative to query count; which is exactly what we need to find the cause.

Installation Copy Chapter

Pre-built Binaries Copy Section

Every release includes checksums.txt with SHA-256 hashes. The executable is self-contained; no adjacent data file is required. Linux -portable archives contain the fully static builds; ordinary platform archives use their native system ABI.

Windows note: SmartScreen may show a warning for unsigned software. Click "More info" → "Run anyway". Verify integrity with checksums.txt.

Setup Scripts Copy Section

Automated download + install

macOS / Linux:

curl -fsSL https://raw.githubusercontent.com/DeusData/codebase-memory-mcp/main/scripts/setup.sh | bash

Windows (PowerShell):

irm https://raw.githubusercontent.com/DeusData/codebase-memory-mcp/main/scripts/setup-windows.ps1 | iex

AUR (Arch Linux) Copy Section

yay -S codebase-memory-mcp-bin

paru -S codebase-memory-mcp-bin

The codebase-memory-mcp-bin package is available at: https://aur.archlinux.org/packages/codebase-memory-mcp-bin

Nix (flake) Copy Section

The flake exposes two server packages plus the standalone frontend:

Run directly without installing:

# Standard server
nix run github:DeusData/codebase-memory-mcp

# Server with CBM Lens embedded
nix run github:DeusData/codebase-memory-mcp#codebase-memory-mcp-ui -- --ui=true --port=9749
# then open http://127.0.0.1:9749

Or build a binary into ./result/bin/codebase-memory-mcp:

nix build github:DeusData/codebase-memory-mcp                          # standard
nix build github:DeusData/codebase-memory-mcp#codebase-memory-mcp-ui   # with UI

Working in a clone? Use . in place of the flake URL, e.g. nix run .#codebase-memory-mcp-ui -- --ui=true, or drop into a shell that puts the binary on PATH with nix shell .#codebase-memory-mcp-ui.

Note: launched by hand (not from an MCP client) the server exits as soon as stdin closes; that's normal MCP behaviour. Keep stdin open while testing the UI, e.g. sleep infinity | codebase-memory-mcp --ui=true --port=9749. The codebase-memory-mcp-ui package embeds the UI at build time; nix run'ing the standard default package with --ui=true will refuse to start the HTTP server.

Install via Claude Code Copy Section

You: "Install this MCP server: https://github.com/DeusData/codebase-memory-mcp"

Build from Source Copy Section

Prerequisites: C/C++ compiler, zlib, and Node.js for CBM Lens

git clone https://github.com/DeusData/codebase-memory-mcp.git
cd codebase-memory-mcp
scripts/build.sh --with-ui          # the shipped composition (CBM Lens embedded, needs Node.js)
scripts/build.sh                    # without the UI (development only)
# Binary at: build/c/codebase-memory-mcp   (codebase-memory-mcp.exe on Windows)

The --with-ui build embeds CBM Lens and the agent integration templates into the native executable. You do not need Node.js to run that executable. Optional browser model weights are downloaded separately after opt-in.

Tests

Counts on main at 4d9592b5, where CBM Lens arrived. Native and Go counts use the macOS configuration; platform-specific registrations differ.

Run the native test suite:

scripts/test.sh                     # full: clean sanitizer build + all suites + guards
scripts/test.sh --suites <name>     # one suite, incremental, seconds
build/c/test-runner --list-suites   # what is available

Frontend development and validation:

cd graph-ui
npm ci
npm run test:unit
npm run build
npm run dev

The dev server listens on port 5173 and proxies /api and /rpc to port 9749. Start the matching CBM backend first.

Package wrappers have their own test commands: go test ./... in pkg/go, npm test in pkg/npm, and python3 -m unittest discover -s tests -p 'test_*.py' in pkg/pypi. The Go wrapper requires Go 1.26.1 or newer.

scripts/test.sh is the C test entry point used by CI. The canonical local artifact-flow check builds both stripped/unstripped candidates, defaults to the stripped candidate for this explicitly unscanned local run, packages those exact bytes, extracts the archive, and smokes it:

scripts/ci/smoke-artifact.sh <linux|darwin|windows> <amd64|arm64>

scripts/package-release.sh is intentionally a lower-level immutable boundary:
it accepts only an already-final --selected-binary plus its
--expected-sha256; it never builds, strips, signs, or relinks the executable.

Manual MCP Configuration Copy Section

If you prefer not to use the install command

Add to ~/.claude.json (user scope) or project .mcp.json:

{
  "mcpServers": {
    "codebase-memory-mcp": {
      "command": "/path/to/codebase-memory-mcp",
      "args": []
    }
  }
}

Restart your agent. Verify with /mcp; you should see codebase-memory-mcp with 17 tools.

Multi-Agent Support Copy Chapter

install configures 45 supported automatic/conditional client surfaces: 39 detected automatically and 6
conditional or explicit. “Conditional” means the installer writes only when the
documented platform or an explicit, already-existing config path proves the
target is active. It never flips experimental feature flags, enables plugins,
YOLO modes, global permission bypasses, or third-party instruction trust.

Where a client has a documented custom-agent format, the installer creates three
exact-owned definitions from one canonical contract:

• Scout (Tier 1): about 3 to 4 narrow calls for fast positive, provisional discovery; no absence, exhaustive-impact, or dead-code claims.
• Verify (Tier 2, default): task-directed graph evidence, exact source checks, path coverage for every cited file, and scope coverage before negative claims.
• Auditor (Tier 3): bounded scope, current index generation, complete relevant pagination, broader relationship checks, and explicit unresolved limitations.

Every direct tier batches check_index_coverage for its evidence paths and reads
flagged ranges or skipped/excluded files directly. A clean coverage result means
only “no recorded gap,” never proof of completeness. Clients without safe child
MCP access receive the same three tiers as parent-handoff agents; the parent must
supply project, generation, pagination state, graph evidence, and coverage
results. Updates migrate only byte-identical prior Verify definitions and never
overwrite user-modified agents.

For Codex, install keeps only a tiny managed activation pointer in global
$CODEX_HOME/AGENTS.md; all detailed behavior lives in the installed codebase-memory skill.
Fresh installs create the pointer, upgrades replace the legacy full managed block while preserving
all user-owned bytes, and uninstall removes only the managed pointer.

Sessions, compaction, and subagents Copy Section

Hooks installed by this project are fail-open and context-only. Claude Code's
PreToolUse observes Grep/Glob/Bash and injects matching graph symbols as
additionalContext; PostToolUse on Read adds targeted coverage context when
the graph could not fully parse or index that file. It never denies or replaces
the requested tool call.

Claude Code, Codex CLI, Qwen Code, GitHub Copilot CLI, and VS Code's Copilot
runtime receive paired session/subagent context where the vendor exposes a
documented context-output contract. Codex users must review and trust installed
hooks through /hooks; changing a hook definition changes its trust hash, so an
update can require re-trust. Qoder uses SessionStart, SubagentStart, and
post-Read coverage, including its documented PowerShell executor on Windows.
Kimi uses UserPromptSubmit, while Hermes uses pre_llm_call; both retain their
documented Windows execution paths. Devin installs
UserPromptSubmit and PostCompaction on macOS/Linux and adds SessionStart
only when Claude's equivalent managed hook is not present. GitLab Duo gets a
narrowly scoped macOS/Linux user SessionStart entry on its experimental hook
surface. GitLab Duo, Devin, and Factory hooks are withheld on Windows
because those vendors do not document a deterministic shell/executor contract
there. Gemini CLI, Factory Droid, and Augment also add documented post-read/view
coverage context but expose no equivalent documented child-start context.

For runtimes without a stable context-producing lifecycle event, durable files
carry the contract across fresh sessions and compaction: verify the graph project
and index freshness, query structural facts in the parent, then pass the project,
qualified symbols, paths, and call-chain evidence in every delegated task.
Claude, Codex, Gemini, Kiro, Qwen, Copilot, CodeBuddy, OpenCode, Kilo, Vibe,
Qoder, Junie, Factory, and Grok Build receive Scout, Verify, and Auditor graph profiles.
Kiro embeds this MCP server with --tool-profile scout for Scout and
--tool-profile analysis for Verify/Auditor. Junie registers equivalent named
server aliases because its subagent schema filters by server rather than by
individual tool. Both process profiles use positive allowlists: Scout exposes
seven fast inspection tools, Analysis exposes eleven, and future or mutating
tools remain unavailable until explicitly reviewed. If either Junie alias
collides with user configuration, the installer preserves it and installs
parent-handoff profiles instead. Qoder combines its documented named-server
selection with exact tier-specific MCP tool IDs. Factory uses exact registered
MCP tool IDs without its additive mcpServers field, which would expose the
whole server. Codex, Kilo, Vibe, and other capable formats likewise enumerate
the narrowest supported tool set. Rovo, Cursor, Augment, Pochi, and Cline use parent handoff where direct
child MCP is unavailable or unsafe; Pochi is limited to readFile, and Cline
child agents cannot use MCP.

Cline's file hooks auto-activate when present, and current Cline does not
reliably consume their context output, so automatic adapters are withheld and
older owned adapters are cleaned up. CodeBuddy's beta, version-gated hooks are
not auto-installed. Junie's EAP
SessionStart output is documented as ignored, so no context hook is installed.
Junie custom agents remain EAP-dependent. Qoder can resolve higher-priority
project or plugin agents before user agents with the same name; reload the
client after installation or profile changes.
Cursor context
hooks are withheld: session context injection has a known race, subagentStart
is control-only, and read-only subagents cannot safely receive MCP access. Grok Build's
passive hook events (SessionStart, SubagentStart, PostToolUse) discard
stdout and PreToolUse honors only deny/rewrite decisions, so its context hooks
are withheld; Grok also reads Claude and Cursor MCP, skill, and hook files
through its compat layer, and the native config.toml entry shadows that copy
by name. Rovo
has no documented session context-output hook, and Bob
documents neither a suitable hook nor a custom-agent surface. Those surfaces are
not approximated with invented augmentation. Kimi plugins, Amp plugins, and
GitLab experimental global skills remain opt-in.

OpenClaw reinjects the Codebase Knowledge Graph (codebase-memory-mcp) AGENTS
section after compaction and places the same guidance in TOOLS.md, the bootstrap
files inherited by its subagents. Automatic augmentation covers the active/default
workspace. Separate agents.list[].workspace directories require making that
workspace active for installation or copying the managed block there.

The installed Claude shim is named cbm-code-discovery-gate for backward
compatibility; despite the legacy name, it never gates or blocks.

Manual or UI-managed integrations Copy Section

These are intentionally not counted as automatic installs: Qodo MCP is added
through its UI and may be governed by enterprise allowlists; Warp MCP is managed
through Warp Drive/UI or per invocation (only the shared skill is automatic);
JetBrains AI Assistant / ACP is IDE-managed; GitHub Copilot coding agent, Jules,
and CodeRabbit are cloud/repository-managed; Replit exposes a remote/service
integration rather than a stable local user-global client; BLACKBOX AI does not
document a stable arbitrary user-global MCP/instruction/agent schema; Plandex has
no stable global registry safe to mutate; and SWE-agent uses explicit YAML and is
no longer a suitable automatic global target.

CLI Mode Copy Chapter

Every MCP tool can be invoked as a local, one-shot command. CLI tools connect to the shared coordination daemon (starting one if none is already running) to get a crash-safe exact-build admission lease for the command's lifetime; they never register with the background watcher, and a daemon a command had to start itself exits again once the command's connection closes and no other session is using it. index_repository is the only exception internally: it starts a temporary, exact-build supervised worker for the index, then stops that worker before the CLI command exits; the worker holds its own lease until exit.

Commands that mutate graph data use shared OS-backed, per-project locks. This serializes conflicting work from CLI and MCP sessions on the same project while allowing unrelated projects to proceed independently.

When stderr is an interactive terminal, the CLI automatically shows lifecycle and indexing progress. Pass --progress to force the same feedback when stderr is redirected or the command is run non-interactively. Pass --quiet to disable automatic terminal progress and ordinary diagnostics while retaining errors; it cannot be combined with --progress or outer cli --verbose. Routine informational logs are quiet by default; pass outer cli --verbose to include them. Progress and logs use stderr while stdout remains reserved for the command result. Read tools return a compact tree by default; pass a tool's --format json for machine-readable payload JSON, or outer --json for the full MCP envelope.

Large compact-tree tables may start with a response-local <section>_refs directory and an explicit <section>_ref_rule. A cell such as @0+handler.go reconstructs to ref 0's prefix plus handler.go. References are local to that sibling <section> table and expansion is non-recursive: entries inside <section>_refs are always literal prefixes. This is limited to declared path and qualified-name columns and activates only when the exact rendered table is at least 15% and 64 bytes smaller and a conservative model-neutral token-shape proxy also improves by at least 1%. Search and trace likewise render direct and prefix-grouped tree shapes and keep the smaller complete representation, so singleton or scattered answers do not pay directory overhead. Keys are declared once per table but never cryptically abbreviated, and --format json keeps stable literal strings for machine consumers. Both gates are deterministic; exact token counts still depend on the caller's tokenizer.

Lean responses truncate semantically, never by cutting arbitrary bytes from code or identifiers. Ranked graph rows are retained ahead of raw grep rows and diagnostic summaries; omitted rows/sections report totals, has_more, and a strictly advancing continuation offset or cursor. If even the first whole row cannot fit, CBM asks for a higher budget and emits no self-looping cursor. max_output_tokens is model-neutral sizing guidance: CBM enforces a deterministic ceiling of four UTF-8 bytes per requested token, so it is not a tokenizer-exact count. Detail flags such as diagnostics, source_mode, and detail opt into heavier fields. search_code pages ranked rows with result_limit/result_offset (limit remains a compatibility alias), raw rows with raw_limit/raw_offset, and directory summaries with directory_limit/directory_offset. Raw lines default to a UTF-8-safe match-centered preview; each row reports content_start_byte, returned/total byte counts, match byte bounds when known, and a content continuation offset. Pass raw_content_offset to page the original line without moving the raw-row cursor. match_limit and source_max_lines bound per-result details, with exact omission metadata. detect_changes pages changed files, impacted symbols, and module summaries independently; prefer its snapshot-bound *_cursor continuations, which reject changed commits, worktree bytes, graph generation, or semantic arguments instead of silently skipping or duplicating rows.

Every response is standard UTF-8. Identifiers, paths, and raw search previews preserve POSIX byte-string identities: a preserved value containing malformed UTF-8 is emitted reversibly as @bytes:<lowercase hex of every original byte>. A valid preserved value that literally begins with the reserved @bytes: or @utf8: prefix is emitted as @utf8:<original value>, so decoding is unambiguous: strip one @utf8: prefix for literal UTF-8, or hex-decode one @bytes: prefix for original bytes. Ordinary valid UTF-8 is unchanged and pays no output-token overhead. To keep code readable, source bodies replace malformed UTF-8 with U+FFFD; use the pageable raw search preview when byte-exact source inspection is required.

Use cli <tool> --help to see the flags generated from that tool's input schema:

codebase-memory-mcp cli index_repository --repo-path /path/to/repo
codebase-memory-mcp cli list_projects

# Use the "name" returned by list_projects as the project value.
codebase-memory-mcp cli search_graph --project my-project --name-pattern '.*Handler.*' --label Function
codebase-memory-mcp cli trace_path --project my-project --function-name Search --direction both
codebase-memory-mcp cli query_graph --project my-project --query 'MATCH (f:Function) RETURN f.name LIMIT 5'

# Force human-readable progress without contaminating stdout.
codebase-memory-mcp cli --progress index_repository --repo-path /path/to/repo
# Suppress automatic terminal progress and non-error diagnostics.
codebase-memory-mcp cli --quiet list_projects --format json
codebase-memory-mcp cli search_graph --project my-project --label Function --format json
codebase-memory-mcp cli list_projects --format json --detail stats | jq '.projects[].name'

Tools requiring arguments accept JSON piped on stdin. Tools with only optional arguments, such as list_projects, accept flags, --args-file, or redirected JSON files without waiting on an inherited pipe. Inline JSON remains accepted for backward compatibility but is deprecated in favor of flags, --args-file, or stdin.

MCP Tools Copy Chapter

Indexing Copy Section

Long indexes and client call deadlines. A synchronous index_repository on a large
repository can take longer than an MCP client allows one tool call (some IDE clients give up
after a fixed deadline). When a client cancels or disconnects, the daemon cancels an index that
nobody else is waiting for, so retrying the same blocking call never finishes. Use the async
mode instead:

1. index_repository(repo_path="/abs/path", async: true) starts the index in the daemon (or
   joins the one already running for that project) and returns immediately with
   state (queued/running). The job keeps running even if the client cancels, times out or
   disconnects; only stopping the daemon ends it.
2. index_repository(repo_path="/abs/path", status: true) reports state
   (queued, running, cancelling, succeeded, failed, cancelled), started_at,
   finished_at and an error summary. Poll it until the state is succeeded, failed or
   cancelled. Pass the same repo_path (and name, if the index call used one).

async and status are exclusive; async does not apply to cross-repo-intelligence. Both
need the daemon-backed server (the default codebase-memory-mcp entry point). A temporary
daemon (started on demand rather than by codebase-memory-mcp daemon start) stops, and
cancels its jobs, when its last client disconnects. A connected MCP session keeps it alive, so
async from your editor works. A one-shot codebase-memory-mcp cli index_repository --async
that is the daemon's only client is refused with a clear error, because the job would die the
moment the command exits: run codebase-memory-mcp daemon start first, keep an MCP session
open, or call without --async. status works everywhere. When a synchronous call was cut
short, the next
index_repository or status call for that project carries a notice suggesting the async
mode. index_status keeps describing the published graph and its freshness.

Querying Copy Section

manage_adr(mode='set_sections') writes one or more sections by name and splices them into the stored document, so text outside the named sections (including a preamble, code fences and section ordering) is preserved byte-for-byte. Any ## Heading works, not just the conventional PURPOSE / STACK / ARCHITECTURE / PATTERNS / TRADEOFFS / PHILOSOPHY set; names match exactly, including case. Writing the same section twice is a no-op, so a retry after a lost response cannot duplicate content.

manage_adr query modes (get and sections) use the server's cached query store so they can proceed while a same-project reindex is running. If another process publishes a replacement store during reindexing, they can return the pre-publication ADR until idle eviction refreshes that cache. Updates remain serialized through the project mutation guard.

Graph Data Model Copy Chapter

Node Labels Copy Section

Project, Package, Folder, File, Module, Section, Class, Struct, Interface, Trait, Enum, Type, Function, Method, Field, Variable, Constant, Macro, Decorator, Route, Channel, Table, Process, EnvVar, Resource, Chart, Image, Branch

Edge Types Copy Section

CONTAINS_FOLDER, CONTAINS_FILE, DEFINES, DEFINES_METHOD, IMPORTS, CALLS, CALL_REFERENCE, USAGE, INHERITS, IMPLEMENTS, OVERRIDE, DECORATES, RAISES, THROWS, WRITES, DATA_FLOWS, HTTP_CALLS, GRPC_CALLS, ASYNC_CALLS, EMITS, LISTENS_ON, HANDLES, SPAWNS, CONFIGURES, INFRA_MAPS, CONNECTS_TO, DEPLOYS, DEPENDS_ON, REFERENCES_FILE, MENTIONS, TESTS, TESTS_FILE, SIMILAR_TO, SEMANTICALLY_RELATED, FILE_CHANGES_WITH, HAS_BRANCH, and the cross-repo CROSS_HTTP_CALLS, CROSS_GRPC_CALLS, CROSS_GRAPHQL_CALLS, CROSS_TRPC_CALLS, CROSS_ASYNC_CALLS, CROSS_CHANNEL

MENTIONS links a documented definition to the code its doc comment references (C# <see cref>, <seealso cref>, <exception cref>, ``), one edge per pair with via, syntax, tier (exact or unique), first line and count. A reference is bound only when it names exactly one definition under the language's own lookup rules; the rest are counted by reason in index_status under doc_links (samples with diagnostics='full').

Qualified Names Copy Section

get_code_snippet uses qualified names: ..<name>. Use search_graph to discover them first.

Supported Cypher (openCypher read subset) Copy Section

query_graph is a read-only openCypher subset:

• Clauses: MATCH, OPTIONAL MATCH, multiple MATCH, WHERE, WITH (+ WITH … WHERE), RETURN, ORDER BY, SKIP, LIMIT, DISTINCT, UNWIND, UNION / UNION ALL, CASE.
• Patterns: labelled nodes, label alternation (n:A|B), relationship types/direction, variable-length paths [*1..3], inline property maps.
• WHERE: = <> < <= > >=, AND/OR/XOR/NOT, IN, CONTAINS, STARTS WITH, ENDS WITH, IS [NOT] NULL, regex =~, label test n:Label, and EXISTS { (n)-[:TYPE]->() } (single-hop existence; great for dead-code, e.g. WHERE NOT EXISTS { (f)<-[:CALLS]-() }).
• Aggregates: count (+DISTINCT), sum, avg, min, max, collect.
• Functions: labels, type, id, keys, properties; toLower/toUpper/toString/toInteger/toFloat/toBoolean; size, length, trim/ltrim/rtrim, reverse; coalesce, substring, replace, left, right.

Anything outside this subset (write/MERGE/CALL clauses, unsupported functions, list/map literals, comprehensions, path functions, parameters) fails with a clear unsupported … error rather than returning empty results.

Ignoring Files Copy Chapter

Layered: hardcoded patterns (.git, node_modules, etc.) → .gitignore hierarchy → .cbmignore (project-specific, gitignore syntax). Symlinks are always skipped.

See docs/cbmignore.md for the full .cbmignore how-to: syntax, precedence across the ignore layers, and negation semantics.

Configuration Copy Chapter

codebase-memory-mcp config list                          # show all settings
codebase-memory-mcp config set auto_index true           # auto-index on session start
codebase-memory-mcp config set auto_index_limit 50000    # max files for auto-index
codebase-memory-mcp config set auto_watch false          # don't register background git watcher (default: true)
codebase-memory-mcp config set doc_candidates false      # skip doc<->code candidate scoring on huge repos (default: auto = on for full, off for moderate/fast)
codebase-memory-mcp config set watcher_enabled false     # stop the watcher thread entirely (default: true)
codebase-memory-mcp config set watch_non_git true        # also poll non-git project roots (default: false)
codebase-memory-mcp config set index_max_files 250000    # optional per-index source-file limit
codebase-memory-mcp config set index_max_source_mb 16384 # optional per-index source-size limit
codebase-memory-mcp config reset auto_index              # reset to default

The two index_max_* settings default to off. Exceeding one fails the complete
index attempt rather than publishing a partial graph; an existing serving index
is preserved. See Index resource limits.

Environment Variables Copy Section

Environment used by daemon-owned components (such as diagnostics, daemon logging, and process-wide indexing resource limits) is captured from the first daemon-backed session that starts the daemon. Later sessions join that process and cannot replace those values. To change them, close all daemon-backed sessions, update the relevant agent configurations consistently, and restart a session. CBM_ALLOWED_ROOT remains session-specific, a conflicting CBM_CACHE_DIR is rejected, and a one-shot CLI command is not exempt from the rule above: it connects to the coordination daemon like any other session, starting one if none is running. Its own environment becomes the captured daemon-owned environment only when its invocation is the one that starts the daemon; joining an already-running daemon, it inherits that daemon's already-captured values instead.

# Store indexes in a custom directory
export CBM_CACHE_DIR=~/my-projects/cbm-data

Custom File Extensions Copy Chapter

The JSON config files support a single key, extra_extensions, which maps additional file extensions to supported languages. Useful for framework-specific extensions like .blade.php (Laravel) or .mjs (ES modules). (For other tunables, see Environment Variables and the config subcommand above.)

Need the full config-file reference? See docs/CONFIGURATION.md.

Per-project (in your repo root):

// .codebase-memory.json
{"extra_extensions": {".blade.php": "php", ".mjs": "javascript"}}

Global (applies to all projects):

// ~/.config/codebase-memory-mcp/config.json  (or $XDG_CONFIG_HOME/...)
{"extra_extensions": {".twig": "html", ".phtml": "php"}}

Each entry maps an extension (which must start with .) to a language name. Language names are matched case-insensitively. Accepted values (aliases in parentheses) are:

arkts, asciidoc, bash (sh), c, c++ (cpp), c# (csharp), chialisp, clojure, cmake, cobol, common lisp (commonlisp, lisp), css, cuda, dart, diagram, dockerfile, elixir, elm, emacs lisp (emacslisp), erlang, f# (fsharp), form, fortran, glsl, go, graphql, groovy, haskell, hcl (terraform), html, ini, java, javascript, jinja2 (jinja), json, julia, kotlin, lean, lua, magma, makefile, markdown, matlab, meson, mojo, nix, objective-c (objc), ocaml, openapi, pdf, perl, php, plsql, protobuf, python, r, ruby, rust, scala, scss, sql, svelte, swift, toml, tsx, typescript, verilog, vimscript, vue, wolfram, xml, yaml, zig.

Project config overrides global for conflicting extensions. An entry whose language name is unknown, or whose extension does not start with ., is skipped and logged as a warning. Missing config files are ignored.

Persistence Copy Chapter

SQLite databases stored at ~/.cache/codebase-memory-mcp/. Persists across restarts (WAL mode, ACID-safe). To reset: rm -rf ~/.cache/codebase-memory-mcp/.

Troubleshooting Copy Chapter

Hybrid LSP Copy Chapter

Semantic type resolution beyond tree-sitter.

Tree-sitter alone gives a syntactic AST. That handles naming, structure, and call sites well, but it can't tell you that user.profile.display_name() resolves to Profile.display_name declared three modules away; tree-sitter doesn't track imports, generics, inheritance, or stdlib types.

codebase-memory-mcp ships a lightweight C implementation of language type-resolution algorithms, structurally inspired by and compatible with major language servers (tsserver / typescript-go, pyright, gopls, Roslyn, Eclipse JDT, rust-analyzer), embedded directly into the native executable. No language server process, no per-project setup, no API key. We call this layer Hybrid LSP: it runs alongside tree-sitter on every parse and refines invocation resolution (CALLS / RESOLVED_CALLS) and callable-value resolution (CALL_REFERENCE, with ambiguous values retained as USAGE) using type information, so the resulting graph mirrors what an IDE "Go to Definition" would resolve.

Languages with full Hybrid LSP:

Two-layer architecture:

1. Tree-sitter pass: fast, syntactic, runs for every one of the 162 languages. Extracts definitions, calls, imports.
2. Hybrid LSP pass: type-aware, runs above the tree-sitter pass per-language. Refines call edges using the import graph plus a per-file or pre-built cross-file definition registry. Languages without a Hybrid LSP pass yet fall back to textual resolution, with coverage varying by language and construct.

The result is a knowledge graph accurate enough to drive trace_path across packages, inheritance hierarchies, and stdlib calls; without paying for a language server process per project.

Language Support Copy Chapter

162 languages, counted by the vendored grammar directories in the parser manifest. Benchmarked against 64 real open-source repositories (78 to 49K nodes):

Also supported (not yet benchmarked): Ada, Agda, Apex, ArkTS, Assembly (NASM), Astro, AWK, Beancount, BibTeX, Bicep, Bitbake, Blade, Cairo, Cap'n Proto, CFML, CFScript, Chialisp, Clojure, CMake, COBOL, Common Lisp, Crystal, CSV, CUDA, D, Devicetree, Diff, .env, Elm, Emacs Lisp, F#, Fennel, Fish, FORM, Fortran, FunC, GDScript, .gitattributes, .gitignore, Gleam, GLSL, GN, Go module, Go template, GraphQL, Hare, HLSL, Hyprlang, INI, ISPC, Janet, Jinja2, JSDoc, JSON, JSON5, Jsonnet, Julia, Just, Kconfig, KDL, Lean 4, Linker Script, Liquid, LLVM IR, Luau, Magma, Makefile, Markdown, MATLAB, Mermaid, Meson, Mojo, Move, Nickel, Nix, ObjectScript (UDL and routines), Odin, Pascal, Pine Script, Pkl, PL/SQL, PO (gettext), Pony, PowerShell, Prisma, .properties, Protobuf, Puppet, PureScript, QML, Racket, Regex, requirements.txt, ReScript, RON, reStructuredText, Scheme, Slang, Smali, Smithy, Solidity, SOQL, SOSL, Squirrel, SSH config, Starlark, Svelte, Sway, SystemVerilog, TableGen, Tcl, Teal, Templ, Thrift, TLA+, Typst, Verilog, VHDL, Vim script, Vue, WGSL, WIT, Wolfram, XML, Zsh.

Architecture Copy Chapter

src/
  main.c              Entry point (MCP stdio server + CLI + install/update/config)
  daemon/             Per-account session coordination, IPC, lifecycle, shared jobs/watchers
  mcp/                MCP server (17 tools, JSON-RPC 2.0, session detection, auto-index)
  cli/                Install/uninstall/update/config (45 client surfaces, hooks, instructions)
  store/              SQLite graph storage (nodes, edges, traversal, search, Louvain)
  pipeline/           Multi-pass indexing (structure → definitions → calls → HTTP links → config → tests)
  cypher/             Cypher query lexer, parser, planner, executor
  discover/           File discovery (.gitignore, .cbmignore, symlink handling)
  watcher/            Background auto-sync (git polling, adaptive intervals)
  traces/             Runtime trace ingestion
  ui/                 Local HTTP server + embedded CBM Lens frontend
  foundation/         Platform abstractions (threads, filesystem, logging, memory)
graph-ui/             CodeAtlasWeb: React, Monaco, and Three.js browser frontend
internal/cbm/         Vendored tree-sitter grammars (162 languages) + AST extraction engine

Security Copy Chapter

Every release is verified through a multi-layer pipeline before publication:

• VirusTotal: all 24 executable candidates (unstripped, debug-stripped and stripped) across the eight release products are scanned before smoke/soak (clean is preferred and only the single-Microsoft !ml tolerance documented in SECURITY.md may pass; the number of engines that returned a decisive result is recorded as evidence but is VirusTotal fleet availability, not a pass condition). The selected executable is packaged without changing its SHA-256, release notes link the verdict for the exact bytes shipped, and the full per-candidate evidence is published alongside the release as TSVs for anyone auditing the selection. Every distinct object extracted from the shipped containers; install.sh, install.ps1, LICENSE, THIRD_PARTY_NOTICES.md, the MCPB manifest.json and the unpacked UI assets; is then scanned as well, so the full published surface is covered, not just the executables.
• SLSA Level 3: cryptographic build provenance generated by the trusted GitHub Actions build workflow; verify with gh attestation verify <file> --repo DeusData/codebase-memory-mcp --signer-workflow DeusData/codebase-memory-mcp/.github/workflows/_build.yml
• Sigstore cosign: keyless signatures on all artifacts; bundles included in every release
• SHA-256 checksums: checksums.txt published with every release; verified by both install scripts before extraction
• CodeQL SAST: blocks release pipeline if any open alerts remain
• No language-runtime dependency chain: the native backend needs no language-runtime environment; UI builds embed the frontend

VirusTotal release scans Copy Section

The v0.11.0 release notes report clean scans for all eight shipped binaries. Each result is tied to the exact executable hash. The links below show the reports for those binaries; the displayed hashes are shortened for readability.

The release also publishes candidate scan results, binary provenance, and selection records. For subsequent versions, use the scan evidence linked from the latest release.

Help shape what comes next Copy Chapter

The next idea could be yours. We invite the OSS community to bring ideas, tricky integration questions, and feedback to our Discord community. When a direct conversation works better, reach out to DeusData or the project's co-developers there. Let's build what comes next together.

License Copy Chapter

MIT

Images README (`docs/images/README.md`) Copy Section

CBM Lens Beta README captures

These are screenshots and browser recordings of the real application using public source from this repository. The application data and text were not retouched. The GIFs were refreshed on 20 October 2026; the still images preserve the earlier captures from 12–13 October.

High-resolution recordings: 20 October Copy Chapter

• Repository: DeusData/codebase-memory-mcp
• Source commit: 4cf44147e639ae070ee44d9230244bfb9d9676e0
• Build: make -f Makefile.cbm cbm-with-ui, with the standard frontend and experimental agent activity disabled
• Capture: built-in browser screenshots on macOS, English locale, native viewport 1600 × 1000, 12 frames per second
• Encoding: JPEG screenshot frames converted to GIF with one 256-color palette per recording, no dithering, infinite looping; no upscaling, sharpening or pointer overlays

The v0.0.1 label is the frontend package version, not the backend release version. Standard native releases include Lens; using a source build for these recordings is a capture-provenance detail, not an installation requirement.

Data and interaction Copy Section

The project codebase-memory-mcp is a clean git archive export of the source commit, indexed in full mode into an isolated local cache. No private repository was used. Because the export has no .git folder, the interface reports "Git unavailable" where applicable.

The index reported 81,190 nodes and 411,072 edges. These counts describe this source scope, not a benchmark or proof of complete repository coverage. The interface retains its coverage and display-limit notices. The clips join recorded interaction segments with pauses shortened between them; their displayed application states and answers are unchanged. Directional pulses represent indexed relationships, not observed execution.

All three GIFs are 1600 × 1000. Their frames and playback duration were decoded and checked after export. README and website use these same files.

Local chat check Copy Section

The optional Qwen2.5 Coder 0.5B model was downloaded and loaded through the application's own setup. Automatic explanations were disabled for this recording. Both displayed answers are deterministic graph answers, labelled "Listed from the indexed graph; not generated by the model." The clip shows one incoming CALLS relationship from cbm_arena_alloc and two outgoing CALLS relationships to cbm_alloc and cbm_free, alongside the other indexed relationship types. It does not demonstrate model-generated prose.

Original still images: 12 October Copy Chapter

Bernhard Jackiewicz captured the PNG stills on 12 October 2026, from commit 4d9592b5 on main after PR #2068. The build used scripts/build.sh --with-ui; headless Chromium on macOS used English locale, a 1440 × 900 viewport and device scale factor 1. Experimental agent activity was disabled.

The indexed project was a public git archive export of that commit in full mode, with 75,037 nodes and 361,349 edges. The PNG stills are 1440 × 900, except the chat still at 1440 × 1200 so the explanation and answer fit together. These are historical application views, also used as reduced-motion alternatives on the website; they are not extracted from the new GIFs.

Refreshed still images: 13 October Copy Chapter

lens-explore.jpg and lens-galaxy.jpg are unedited browser captures at 1280 × 720 from the locally served CBM Lens Beta UI. The running binary reports dev; its exact build commit is unavailable, so these captures do not certify an exact source revision. The local project is named cbm-triage-main-20260919; the screenshots show public CBM source and indexed relationships.

• Explore: src/foundation/limits.c, its compact file impact summary and direct dependencies (24 nodes, 26 edges). The UI's freshness, sampling and coverage notices remain visible.
• Galaxy: cbm_max_file_bytes, both directions, all edge types, two layers (160 nodes, 218 edges).

These counts describe the displayed index snapshot, not a benchmark or proof of complete repository coverage.

BENCHMARK (`docs/BENCHMARK.md`) Copy Section

Codebase Memory MCP -- v0.3.0 Language Benchmark

Methodology Copy Chapter

• 63 languages (27 programming + 8 config/markup), 12 questions each (4 for config languages)
• Up to 5 attempts per question with escalating retry strategies
• Real open-source repos (medium to large: 78--49K nodes)
• Grading: PASS (1.0) / PARTIAL (0.5) / FAIL (0.0), N/A excluded from denominator
• Platform: Apple M3 Pro, macOS Darwin 25.3.0
• Date: 2026-03-01

The 12 Questions Copy Section

Config/markup languages (HTML, CSS, SCSS, YAML, TOML, HCL, SQL, Dockerfile) run Q1, Q4, Q6, Q12. Other questions marked N/A.

Summary Table Copy Chapter

Tier Classification Copy Chapter

Per-Language Results Copy Chapter

Python (django/django) Copy Section

Project: django-python | Repo: /tmp/lang-bench/django-python
Nodes: 49,398 | Edges: 196,022

Score: 10.5/12 (87%)

JavaScript (lodash/lodash) Copy Section

Project: lodash-js | Repo: /tmp/lang-bench/lodash-js
Nodes: 244 | Edges: 405

Score: 9.5/11 (86%)

TypeScript (nestjs/nest) Copy Section

Project: nestjs-ts | Repo: /tmp/lang-bench/nestjs-ts
Nodes: 9,063 | Edges: 15,772

Score: 10.5/12 (87%)

TSX (shadcn-ui/ui) Copy Section

Project: shadcn-tsx | Repo: /tmp/lang-bench/shadcn-tsx
Nodes: 29,755 | Edges: 41,883

Score: 10.5/12 (87%)

Go (codebase-memory-mcp) Copy Section

Project: codebase-memory-mcp | Repo: self
Nodes: 2,259 | Edges: 6,561

Score: 10.5/12 (87%)

Rust (BurntSushi/ripgrep) Copy Section

Project: ripgrep-rust | Repo: /tmp/lang-bench/ripgrep-rust
Nodes: 4,118 | Edges: 6,971

Score: 10.5/12 (87%)

Java (spring-projects/spring-petclinic) Copy Section

Project: spring-petclinic-java | Repo: /tmp/lang-bench/spring-petclinic-java
Nodes: 660 | Edges: 1,080

Score: 10.5/12 (87%)

C++ (nlohmann/json) Copy Section

Project: nlohmann-json-cpp | Repo: /tmp/lang-bench/nlohmann-json-cpp
Nodes: 5,262 | Edges: 8,681

Score: 12/12 (100%)

C# (jasontaylordev/CleanArchitecture) Copy Section

Project: clean-architecture-csharp | Repo: /tmp/lang-bench/clean-architecture-csharp
Nodes: 1,043 | Edges: 1,632

Score: 9/12 (75%)

PHP (laravel/framework) Copy Section

Project: laravel-framework-php | Repo: /tmp/lang-bench/laravel-framework-php
Nodes: 39,767 | Edges: 148,440 (post-Phase-4–5 PHP-LSP; was 196,979 baseline, –25%)
CALLS edges: 47,341 (was ~83k baseline — 43% reduction in name-fallback misroutes)
Tests: 278 PHP-LSP unit tests, all passing (total project: 3,091 / 0 failed)
LSP module size: ~3,700 lines C resolver + ~700 lines stdlib + ~5,500 lines tests = ~9,900 LoC

PHP runs through a Light Semantic Pass (internal/cbm/lsp/php_lsp.c,
~3,500 lines) that approaches phpactor-grade type resolution while
staying in-process and PHP-runtime-free. Phase 4 capabilities:

• Receiver-type tracking with full ancestor chain walk (cycle-detected,
  bounded to 32 hops)
• Namespace + use-clause resolution (class, function, const), incl.
  as aliasing in both wrapped and bare forms
• PHPDoc @var, @param, @property, @method parsing — generics
  (Collection, array) supported
• Type narrowing: instanceof, is_string/int/array/..., assert(...)
  sequential narrowing, negative narrowing on early return / throw
• Property type tracking: typed declarations, constructor property
  promotion, constructor-body inference ($this->bar = $bar)
• Trait flattening with as aliasing
• Late static binding chain depth, self/parent/static distinction
• Match / ternary / clone / cast result-type evaluation
• Foreach element-type propagation through array<T> and Iterator
• Magic methods (__call/__callStatic → facade dispatch)
• Stdlib coverage: SPL, PSR, DateTime, Throwable, Closure, plus
  Eloquent Builder/Model/Collection chains, Symfony HttpFoundation,
  Carbon date methods, PSR-7 with-builders

Attribution correctness fix (the primary metric, see
docs/PHP_LSP_PRE_FLIGHT.md §6):

• ConfiguresPrompts.configurePrompts → helpers.value misroute (3
  call sites): fixed — those typed-receiver $prompt->value() calls
  no longer route to the global helper. Verified: 0 misroute edges.
• Roughly 27 500 fewer total edges, primarily from blocking
  name-fallback edges where the receiver was a vendor (un-indexed) type
  — better precision at the cost of some recall, see notes below.

Score: 10/12 (83%) — the two remaining PARTIALs are non-LSP issues
tracked separately. PHP-LSP delivered the underlying graph-correctness
win (collide-set attribution); benchmark Tier 1 (≥ 90 %) requires the
search-recall and trace-disambiguation tickets to ship as well.

Note on edge-count drop: when the LSP knows the receiver is typed
but the receiver class is not indexed (e.g., Composer vendor types like
Laravel\\Prompts\\Prompt), it emits a synthetic
php_method_typed_unindexed resolution that the bridge uses to
suppress the unified extractor's name-based fallback. This trades
recall (no edge instead of a guess) for precision (no wrong edge). The
pre-flight argued this is the right trade for graph correctness.

Test coverage: 100 unit tests in tests/test_php_lsp.c, all
passing. Total project tests: 2913 / 0 failed.

Lua (neovim/neovim) Copy Section

Project: neovim-lua | Repo: /tmp/lang-bench/neovim-lua
Nodes: 23,955 | Edges: 90,247

Score: 12/12 (100%)

Scala (playframework/playframework) Copy Section

Project: playframework-scala | Repo: /tmp/lang-bench/playframework-scala
Nodes: 19,627 | Edges: 43,764

Score: 9/12 (75%)

Kotlin (ktorio/ktor) Copy Section

Project: ktor-kotlin | Repo: /tmp/lang-bench/ktor-kotlin
Nodes: 25,297 | Edges: 71,498

Score: 12/12 (100%)

Ruby (sinatra/sinatra) Copy Section

Project: sinatra-ruby | Repo: /tmp/lang-bench/sinatra-ruby
Nodes: 1,377 | Edges: 1,893

Score: 9/12 (75%)

C (jqlang/jq) Copy Section

Project: jq-c | Repo: /tmp/lang-bench/jq-c
Nodes: 1,330 | Edges: 3,923

Score: 11/11 (100%)

Bash (bats-core/bats-core) Copy Section

Project: bats-bash | Repo: /tmp/lang-bench/bats-bash
Nodes: 436 | Edges: 479

Score: 10/10 (100%)

Zig (zigtools/zls) Copy Section

Project: zls-zig | Repo: /tmp/lang-bench/zls-zig
Nodes: 2,824 | Edges: 10,230

Score: 10/10 (100%)

Elixir (elixir-plug/plug) Copy Section

Project: plug-elixir | Repo: /tmp/lang-bench/plug-elixir
Nodes: 870 | Edges: 865

Score: 9.5/11 (86%)

Haskell (PostgREST/postgrest) Copy Section

Project: postgrest-haskell | Repo: /tmp/lang-bench/postgrest-haskell
Nodes: 2,066 | Edges: 2,463

Score: 7.5/12 (62%)

OCaml (ocaml/dune) Copy Section

Project: dune-ocaml | Repo: /tmp/lang-bench/dune-ocaml
Nodes: 11,691 | Edges: 10,447

Score: 8/11 (72%)

Erlang (ninenines/cowboy) Copy Section

Project: cowboy-erlang | Repo: /tmp/lang-bench/cowboy-erlang
Nodes: 3,270 | Edges: 9,815

Score: 9.5/11 (86%)

R (tidyverse/dplyr) Copy Section

Project: dplyr-r | Repo: /tmp/lang-bench/dplyr-r
Nodes: 1,618 | Edges: 2,409

Score: 10.5/12 (87%)

Objective-C (AFNetworking/AFNetworking) Copy Section

Project: afnetworking-objc | Repo: /tmp/lang-bench/afnetworking-objc
Nodes: 1,087 | Edges: 1,348

Score: 12/12 (100%)

Swift (Alamofire/Alamofire) Copy Section

Project: alamofire-swift | Repo: /tmp/lang-bench/alamofire-swift
Nodes: 3,631 | Edges: 7,639

Score: 10.5/11 (95%)

Dart (felangel/bloc) Copy Section

Project: bloc-dart | Repo: /tmp/lang-bench/bloc-dart
Nodes: 5,089 | Edges: 6,430

Score: 10.5/12 (87%)

Perl (mojolicious/mojo) Copy Section

Project: mojo-perl | Repo: /tmp/lang-bench/mojo-perl
Nodes: 3,287 | Edges: 4,182

Score: 12/12 (100%)

Groovy (spockframework/spock) Copy Section

Project: spock-groovy | Repo: /tmp/lang-bench/spock-groovy
Nodes: 14,081 | Edges: 34,557

Score: 12/12 (100%)

SQL (flyway/flyway) Copy Section

Project: flyway-sql | Repo: /tmp/lang-bench/flyway-sql
Nodes: 10,149 | Edges: 23,222

Score: 4.5/6 (75%)

Dockerfile (docker-library/official-images) Copy Section

Project: official-images-dockerfile | Repo: /tmp/lang-bench/official-images-dockerfile
Nodes: 1,481 | Edges: 1,588

Score: 4/4 (100%)

CSS (animate-css/animate.css) Copy Section

Project: animate-css | Repo: /tmp/lang-bench/animate-css
Nodes: 295 | Edges: 214

Score: 4/4 (100%)

YAML (kubernetes/examples) Copy Section

Project: k8s-yaml | Repo: /tmp/lang-bench/k8s-yaml
Nodes: 2,110 | Edges: 1,844

Score: 4/4 (100%)

TOML (rust-lang/cargo) Copy Section

Project: cargo-toml | Repo: /tmp/lang-bench/cargo-toml
Nodes: 16,773 | Edges: 51,403

Score: 4/4 (100%)

HCL (hashicorp/terraform) Copy Section

Project: terraform-hcl | Repo: /tmp/lang-bench/terraform-hcl
Nodes: 78 | Edges: 76

Score: 4/4 (100%)

HTML (twbs/bootstrap) Copy Section

Project: bootstrap-scss (shared) | Repo: /tmp/lang-bench/bootstrap-scss
Nodes: 2,726 | Edges: 4,135

Score: 4/4 (100%)

SCSS (twbs/bootstrap) Copy Section

Project: bootstrap-scss (shared) | Repo: /tmp/lang-bench/bootstrap-scss
Nodes: 2,726 | Edges: 4,135

Score: 4/4 (100%)

Linux Kernel Stress Test Copy Chapter

Project: linux-kernel | Subset: drivers/net/ethernet/intel/
Files: 387 C/H files | Nodes: 19,993 | Edges: 67,305

Index Breakdown Copy Section

Results Copy Section

Score: 11/11 (100%) (Q11 N/A)

Stress Metrics Copy Section

The tool handles 20K-node kernel subsystems without timeouts. Deep traces on well-connected probe functions (out_degree=59) produce massive but complete results. Community detection identified 734 clusters -- useful for understanding driver module boundaries.

Cross-Cutting Findings Copy Chapter

Strengths Copy Section

1. Zero indexing failures across 63 repos of all sizes (78 to 49K nodes)
2. 100% on 17 languages -- half the languages achieved perfect scores
3. No language below 62% -- even the weakest performs core operations
4. Massive codebases handled: Django (49K nodes), Laravel (38K nodes), neovim (24K nodes) -- no performance issues
5. Kernel-scale stress test passed: 20K nodes, 67K edges, 129K-char traces, zero timeouts
6. "Removed" languages still strong: Obj-C (100%), Perl (100%), Groovy (100%), Swift (95%), Dart (87%)

Common PARTIAL Patterns Copy Section

• Q10 (Properties): Most languages return functions with properties=null. Parameter/return extraction is incomplete -- this is the single most common deduction
• Q7/Q8 (Trace): Abstract methods, entry-point functions, and built-in references correctly return 0 edges (no false positives), but are graded PARTIAL when no alternative function has edges
• Q11 (Inheritance): Languages using alternative paradigms (traits, protocols, typeclasses, modules) have limited INHERITS edges

Known Limitations Copy Section

1. Function properties: Parameter types and return types are not extracted for most languages, resulting in properties=null
2. Haskell call tracing: Function composition (f . g . h) is not modeled as CALLS edges. Only explicit function applications are traced
3. OCaml call analysis: Limited call resolution due to module functor indirection
4. Cypher IS NOT NULL: Not supported in the Cypher parser -- use alternative query patterns
5. Scala/Ruby trace: Abstract methods and block-based patterns can have 0 trace edges
6. query_graph 200-row cap: Aggregation queries (COUNT) silently undercount on large codebases

Aggregate Statistics Copy Chapter

CONFIGURATION (`docs/CONFIGURATION.md`) Copy Section

Configuration Reference

This page documents the configuration files that codebase-memory-mcp reads or writes today.

At a Glance Copy Chapter

CBM resolves CBM_CACHE_DIR to a canonical per-account path before using any of these locations. The log directory and files are private to the account.

1. Custom File Extension Mapping Copy Chapter

Two optional JSON files let you map additional file extensions to built-in languages.

Global config Copy Section

Default path:

$XDG_CONFIG_HOME/codebase-memory-mcp/config.json

Fallback when XDG_CONFIG_HOME is unset:

~/.config/codebase-memory-mcp/config.json

Per-project config Copy Section

Place this file in the repository root:

.codebase-memory.json

Format Copy Section

{
  "extra_extensions": {
    ".blade.php": "php",
    ".mjs": "javascript",
    ".twig": "html"
  }
}

Notes:

• Extension keys must include the leading dot.
• Language names are case-insensitive.
• Unknown language names are skipped.
• Missing files are ignored.
• If the same extension appears in both files, the per-project file wins.

2. CLI-Managed Runtime Settings Copy Chapter

The config subcommand stores runtime settings in a small SQLite database:

${CBM_CACHE_DIR:-~/.cache/codebase-memory-mcp}/_config.db

Inspect or change values with the CLI:

codebase-memory-mcp config list
codebase-memory-mcp config get auto_index
codebase-memory-mcp config set auto_index true
codebase-memory-mcp config set auto_index_limit 50000
codebase-memory-mcp config set watcher_enabled false
codebase-memory-mcp config reset auto_index

Current keys:

For a watched project, index_status names the polling method in watch.strategy:
git (HEAD and dirty-state polling), tree (a non-git root polled by the
watch_non_git file-tree scan), none (a non-git root that is registered but never
polled, the default for a plain directory) or pending (registered; the first poll
has not run yet).

watcher_enabled vs auto_watch. watcher_enabled controls whether the
watcher subsystem starts at all (the background poll thread). auto_watch is
narrower: it only controls whether a connecting session registers its own
project with an already-running watcher. When watcher_enabled=false,
auto_watch has no effect — there is no watcher to register with.

They also differ in when they are read, which matters because the watcher
lives in the background daemon, not in your MCP client:

• auto_watch is consulted each time a session would register its project, so
  a change applies to sessions that connect afterwards.
• watcher_enabled is read once, when the daemon starts, because it decides
  whether the watcher is built at all. The daemon is long-lived and outlives
  individual MCP sessions, so reconnecting your client is not enough — retire
  the daemon so the next one picks the new value up:

bash Copy

codebase-memory-mcp config set watcher_enabled false
codebase-memory-mcp daemon stop     # next session starts a daemon without the watcher
codebase-memory-mcp daemon status   # confirm

Disabling the watcher does not disable anything else: the daemon still starts,
auto_index still runs, and index_repository stays available for manual
reindexing.

The two index_max_* settings are independent and disabled by default. They
apply to explicit indexing, automatic indexing, and watcher re-indexing, but not
to cross-repo-intelligence, which does not scan repository source files.
Equality is allowed; exceeding either setting fails the complete index request
and preserves any previously serving database. See
Index resource limits for counting, validation, and
error-response details.

3. UI Settings Copy Chapter

The optional built-in graph UI stores its settings in:

${CBM_CACHE_DIR:-~/.cache/codebase-memory-mcp}/config.json

Current format:

{
  "ui_enabled": false,
  "ui_port": 9749
}

Notes:

• If a UI-enabled binary finds its verified external asset pack and no UI config file exists yet, the UI auto-enables on first run. Missing or invalid assets leave the MCP/daemon service available and keep the UI disabled.
• CBM_CACHE_DIR changes both the UI config location and the runtime settings database location.
• CBM resolves CBM_CACHE_DIR to one canonical per-account cache root. A process configured with a different root fails while any CBM session or command is active; close them before switching roots.

4. Environment Variables Copy Chapter

These environment variables affect runtime behavior:

Relocating the daemon rendezvous directory Copy Section

Before it is used, the rendezvous directory and every ancestor of it are checked:
each ancestor must be owned by you or by root, must not be world-writable (unless
it is the standard root-owned sticky directory such as /tmp), and must carry no
allow-ACL — on Windows, no ACE granting mutation rights to another identity. The
rendezvous directory itself is then forced to owner-only (0700, no extended ACL
/ an owner-only DACL).

That ancestry is not always acceptable in the default location. A Windows profile
that has acquired a capability-SID ACE with WRITE_DAC / WRITE_OWNER / DELETE
on %LOCALAPPDATA% — something an installed packaged app can add — fails the walk,
and so can an unusual /tmp or home directory on POSIX. When that happens every
command fails, config list included, so the settings surface cannot be reached
either:

codebase-memory-mcp: secure daemon endpoint could not be created

CBM_RUNTIME_DIR points the rendezvous at an ancestry you choose:

export CBM_RUNTIME_DIR="$HOME/cbm-runtime"   # any directory you own

$env:CBM_RUNTIME_DIR = "D:\cbm-runtime"

The check is not relaxed for the directory you name: it goes through exactly the
same validation as the default, and a value that fails it is refused rather than
silently ignored. Because the rendezvous is how sessions find each other, every
process that should share one daemon must see the same value — set it in the
environment of your MCP client and your shell alike, or a CLI invocation without
it will coordinate through the default location instead.

WSL2 and Windows drives (/mnt/c, /mnt/e, ...). CBM_CACHE_DIR goes through
the same private-directory check. With WSL's default automount options, DrvFs
reports every directory as 0777 and ignores chmod, so a cache under
/mnt/<drive> is refused, and the refusal names this remedy. Either turn on
permission metadata in /etc/wsl.conf, then run wsl --shutdown and reopen WSL:

[automount]
options = "metadata,umask=22,fmask=11"

or keep the cache on the Linux filesystem (the default ~/.cache/codebase-memory-mcp),
which is also much faster than a 9p-mounted Windows drive.

Environment used by daemon-owned components—such as diagnostics, daemon logging, and process-wide indexing resource limits—is captured from the first daemon-backed session that starts the daemon. Later sessions join the existing process and cannot replace those values. To change them, close every daemon-backed session, update the relevant agent configurations consistently, and restart a session. CBM_ALLOWED_ROOT remains session-specific, a conflicting CBM_CACHE_DIR is rejected, and a one-shot CLI command is not exempt from the rule above: it connects to the coordination daemon like any other session, starting one if none is running, so its own environment becomes the captured daemon-owned environment only when its invocation is the one that starts the daemon — joining an already-running daemon, it inherits that daemon's already-captured values instead.

Roots that are always refused Copy Section

Independently of CBM_ALLOWED_ROOT, some directories are refused as an indexing
root because they are too broad or too sensitive to index as a unit:

• a filesystem root, a Windows drive root, or a UNC share root;
• a top-level system tree — /etc, /var, /usr, /home, /Users, and on
  Windows C:\Windows, C:\Users, C:\ProgramData, C:\Program Files;
• your home directory itself (directories below it are fine);
• a credential directory at any depth — .ssh, .aws, .gnupg, .kube,
  .docker, .netrc, .git-credentials, .password-store, macOS Keychains.

Two limits are worth stating plainly. This constrains scope, not
sensitivity: inside a root that is allowed, every file the process can read may
be indexed and later returned. And the credential list is a denylist, so it
raises the cost of a mistake rather than closing the class — a directory it does
not name is permitted.

Shared / HPC filesystems Copy Section

On clusters and other shared machines, home directories often live on a network
filesystem (NFS, Lustre, GPFS, SMB/CIFS), and ~/.cache is sometimes a symlink
into a shared tree. Three rules keep CBM working there:

• Put the cache and the rendezvous on local storage. Set CBM_CACHE_DIR (indexes,
  _config.db) and, if needed, CBM_RUNTIME_DIR (the daemon rendezvous) to a
  local disk you own, for example /local/scratch/$USER/cbm or /tmp/$USER-cbm.
  Set the same values in your MCP client's environment and your shell.

• Do not route the cache into another account's tree. A symlinked ~/.cache is
  resolved and allowed, but every directory on the resolved path must be owned
  by you or by root. If a parent such as /shared/fs belongs to another account,
  that account can rename or replace anything below it, so CBM refuses the path
  and names the directory and its owner:

  text Copy

  ... (cache-private) - '/shared/fs' is owned by uid 1234 (alice), not by you (uid 1000) or root; ...

  This check has no override. Choose a location with a safe ancestry instead.

• Avoid network filesystems for the cache. The indexes are SQLite databases in
  WAL mode, which needs shared-memory locking that network filesystems do not
  provide reliably. On Linux, CBM logs one daemon.cache_root_network_fs warning
  at startup when the cache root is on NFS, Lustre, GPFS, CIFS or SMB2. The cache
  still works there, but it can be slow, and a database can be corrupted.

5. Agent and Editor Integration Files Copy Chapter

The install command can also write MCP entries and instruction blocks into agent/editor config files such as Claude Code, Codex, Gemini, VS Code, Cursor, Zed, and others.

Those target paths vary by tool and platform, so the easiest way to inspect the exact files for your machine is:

codebase-memory-mcp install --dry-run

That prints the specific config files the installer would modify without writing anything.

INDEX RESOURCE LIMITS (`docs/INDEX_RESOURCE_LIMITS.md`) Copy Section

Index resource limits

Index resource limits are optional operator controls for repositories whose
discovery breadth is not known in advance. They are disabled by default so
existing large-repository workloads retain their current behavior.

Discovery settings Copy Chapter