cartograph
Semantic code intelligence for AI coding agents — 45% fewer tool calls, 17% faster exploration, 100% local. A fork of codegraph.
Install / Use
claude mcp add adder-factory -- npx -y github:adder-factory/cartographIf the server publishes to npm under a different name, use that package instead — check the repo README.
MCP Server
Model Context Protocol server
Quality Score
Category
Data & AnalyticsSupported Platforms
Our assessment of cartograph
cartograph scores 83/100 on our quality scale, 481st of 599 Data & Analytics skills we index.
Its MCP Server is 20 KB long, well organised into 22 sections with 13 code examples: a thorough specification that gives an agent plenty to work with.
It has 3 GitHub stars, so there is little community track record yet; judge it on its content.
Maintenance, license and trust
- The repository was last updated 4 days ago, so cartograph is actively maintained.
- It is released under the MIT license, a permissive license that allows use, modification and commercial use with attribution.
- Its trust signals score 92/100, with 1 caution from licensing, adoption, age or documentation. These come from repository metadata, not a code audit — read the skill file before letting an agent act on it.
Safety scan
No issues foundOur scan of the whole file found no instruction hijacking, hidden characters, credential access, data exfiltration or destructive commands (2 minor notes below). An AI review of the same text found nothing harmful.
- noteInstalls by piping a downloaded script into a shellline 78
curl -fsSL https://raw.githubusercontent.com/adder-factory/cartograph/main/install.sh | sh - noteInstalls by piping a downloaded script into a shellline 85
irm https://raw.githubusercontent.com/adder-factory/cartograph/main/install.ps1 | iex
AI review by kimi-k2.7-code on 2026-10-31. Automated pattern scan on 2026-10-31. It catches known dangerous patterns, not every risk — read a skill before letting an agent act on it.
cartograph compared with similar skills
All 4 of these similar skills score higher than cartograph; compare them before choosing.
| Skill | Score | Stars | Updated | Format |
|---|---|---|---|---|
| cartograph (this skill)by adder-factory | 83 | 3 | 4d ago | MCP Server |
| Agent-Reachby Panniantong | 100 | 107.1k | 6d ago | CLAUDE.md |
| claude-memby thedotmack | 100 | 107.0k | today | CLAUDE.md |
| headroomby headroomlabs-ai | 100 | 78.4k | today | CLAUDE.md |
| CowAgentby zhayujie | 100 | 47.7k | today | CLAUDE.md |
Frequently asked questions
- How do I install cartograph?
- Run
claude mcp add adder-factory -- npx -y github:adder-factory/cartograph. The install tabs above show the steps for each supported agent. - Which AI agents does cartograph work with?
- It is written for Claude Code and Claude Desktop, as a MCP Server file. Other agents that read the same format can often use it too.
- Is cartograph safe to use?
- Our scan of the whole file found no instruction hijacking, hidden characters, credential access, data exfiltration or destructive commands (2 minor notes below). An AI review of the same text found nothing harmful. It is MIT-licensed and scores 92/100 on trust signals. Skills are instructions an agent will follow, so read the file before installing it and do not approve commands you do not understand.
- Is cartograph still maintained?
- The repository was last updated 4 days ago, so cartograph is actively maintained.
Skill content
View source on GitHubCartograph
Cartograph 3.0 — current code intelligence, stored with your project.
Native Rust CLI and MCP server · per-project RocksDB · code-aware BM25 · typed code graph
Quick start · Agent workflow · Languages · Architecture · Documentation · Changelog
</div>Cartograph turns a source checkout into a searchable, immutable-generation code graph. It gives coding agents compact evidence about declarations, references, call flow, change impact, affected tests, and source freshness through one native CLI and the Model Context Protocol.
The source checkout remains the source of truth. Every evidence packet carries generation provenance, freshness, confidence, truncation, and explicit abstention instead of presenting stale or incomplete data as certainty.
[!IMPORTANT] Version 3.0 replaces PostgreSQL with embedded RocksDB. Each project owns its database, and a persistent native service keeps its graph current. Existing v2 projects need a source rebuild and host restart; see Upgrade.
Start here
| I want to… | Go to | | --- | --- | | Install Cartograph and build a first graph | Quick start | | Hand setup to a coding agent | Agent-assisted installation | | Connect Codex, Claude Code, Cursor, or another host | MCP usage | | Look up a command, flag, or JSON field | CLI reference | | Check a language, extension, or framework | Language support matrix | | Inspect and maintain native storage | Storage and operations | | Fix a failed setup, error code, or stale graph | Troubleshooting | | See what changed in each release | Changelog | | Browse every guide by task | Documentation home |
What Cartograph gives an agent
| Question | Evidence | | --- | --- | | Where is this declared or referenced? | Exact symbol, path, reference, and identifier lookup | | Which code is most relevant? | Code-aware BM25 over names, implementation identifiers, and documentation | | What calls this, and what does it call? | Typed callers, callees, imports, references, and shortest paths | | What could this change affect? | Bounded reverse impact and structurally connected tests | | Where is numerical behavior risky or statically unknown? | Generation-scoped Rust operation/hazard/precision sites with exact spans and explicit evidence limits | | Is the graph current? | Immutable generation identity and exact supported-source freshness | | What changed in the working tree? | Separately labeled live overlay and Git-ref review packets | | Is an LLM required? | No for exact, lexical, graph, review, freshness, or affected-test workflows |
Cartograph also supports standard SCIP export, persistent per-file SCIP overlays, model-scoped semantic retrieval, generated artifacts with explicit provenance, and deterministic task-intent routing.
Quick start
[!TIP] Prefer agent-assisted setup? Give your coding agent the copy-paste task in Agent-assisted installation.
1. Install the native executable
macOS (Apple Silicon, macOS 26+) or Linux (ARM64/x64, glibc 2.41+):
curl -fsSL https://raw.githubusercontent.com/adder-factory/cartograph/main/install.sh | sh
cartograph --version
Windows (x64), in PowerShell:
irm https://raw.githubusercontent.com/adder-factory/cartograph/main/install.ps1 | iex
cartograph --version
The installer verifies the archive against the release's SHA256SUMS. To build
from source, use the pinned Rust toolchain, a C/C++ compiler, and libclang:
cargo build --locked --release -p cartograph-cli
2. Build the first graph
cd /path/to/project
cartograph index .
cartograph doctor . --format json
cartograph status . --format json
cartograph context 'explain the primary request flow' --project-path .
Each project owns .cartograph/data; no server or Docker bootstrap is needed.
An unchanged second index skips publication. Require a fresh generation and
real retrieval evidence before calling setup ready. Doctor reports native
capabilities separately from freshness, models, registration, and transport.
For a previously indexed PostgreSQL project, run
cartograph db migrate --to rocksdb --project-path . to rebuild and verify the
native destination while preserving legacy data. See the
migration guide for the scope of that rebuild.
3. Connect a coding agent
Cartograph writes project-scoped MCP configuration and pins the absolute native
executable path. For Claude Code the project entry lives under
projects.<path>.mcpServers in ~/.claude.json, Claude's local scope.
cartograph install --yes --target codex --location local --project-path . # OpenAI Codex
cartograph install --yes --target claude --location local --project-path . # Claude Code
cartograph install --yes --target cursor --location local --project-path . # Cursor
The installer supports 19 hosts in total; see the CLI reference for every target ID.
4. Verify the live integration
A setup is ready only after all four signals pass:
cartograph doctorproves database capabilities.cartograph indexpublishes one complete generation.cartograph statusreports that generation as fresh.- A real
findorcontextquery returns generation-scoped evidence.
A CLI request proves the executable and database path. After restarting an
agent host, make one live MCP request as the separate transport-health check.
Until that happens, doctor deliberately reports liveTransport as
not_checked; an on-disk registration never proves which child process an
already-running host loaded.
Keep the graph current
On macOS, start a project service that stays alive when an MCP client disconnects:
cartograph service start --project-path . --format json
cartograph service status --project-path . --format json
Filesystem events use a 150 ms quiet window and incremental file parsing.
Continuous edits are coalesced for at most two seconds, and a periodic source
check catches missed events. These are scheduling bounds; end-to-end freshness
also includes extraction, publication, and the query. Check autoSync and
generation freshness instead of assuming every repository updates in a second.
On Linux, run cartograph watch --project-path . under your user service
manager, or in the foreground. Automatic service registration currently uses
macOS launchd. On Windows, the serving MCP primary owns the watcher.
Upgrade
For an existing v2 installation, install v3 with the installer above, then run the v3 executable:
cartograph db migrate --to rocksdb --project-path . --format json
cartograph upgrade --apply --project-path . --json
Migration rebuilds the graph from current source and verifies retrieval before
recording completion. It preserves the old PostgreSQL database. Legacy notes,
sessions, embeddings, and artifacts are not imported; keep the old data if you
need them. The upgrade repairs recognized host pins and reports any remaining
manual steps. Reopen attached coding hosts to load v3, then verify live MCP
status and a real query. Future native upgrades use the same upgrade --apply
command after stopping the existing owner. For a macOS native service:
cartograph service stop --project-path . --format json
cartograph upgrade --apply --project-path . --json
cartograph service start --project-path . --format json
cartograph status . --format json
For a manually managed owner, stop that project's watch or serving process
cooperatively, upgrade, then start it again through its service manager. Start
the new owner before reconnecting MCP hosts. Verify the status report's version
as well as the client's --version; a new client can still proxy to an older
owner. See the migration guide.
Platform and database support
RocksDB is compiled into the Rust executable on macOS, Linux, and Windows. One primary owns each project's database. macOS/Linux clients can reuse a running MCP owner through private same-user Unix IPC. Windows currently requires closing the existing primary before another process opens that project. PostgreSQL connection flags and managed-container operations are retired.
See storage operations for concurrency, maintenance, source-rebuild migration, and current capacity limits.
Agent workflow
A reliable coding loop starts with freshness, narrows with structural evidence, and closes with impact-aware verification:
cartograph status . # 1. is the graph fresh?
cartograph context 'fix authentication token validation' --project-path . # 2. task evidence
cartograph find 'validateToken' --by name --project-path . # 3. exact anchor
cartograph graph <symbol-id> --direction impact --project-path . # 4. blast radius
cartograph affected --symbol-id <symbol-id> --project-path . # 5. tests to run
cartograph review --ref main --project-path . # 6. review the diff
If status is stale, run cartograph index . or use the bounded MCP admin action.
Indexing unchanged source is a no-op; changed source publishes a complete new
generation atomically.
Core MCP tools
| Tool | Purpose |
| --- | --- |
| cartograph_status | Current generation, counts, capability readiness, and freshness |
| cartograph_find | Exact name/path/reference lookup or code-aware BM25 |
| cartograph_context | Intent-aware evidence, graph context, and edit candidates |
| cartograph_entry_points | Routes, commands, MCP tools, exports, and API boundaries |
| cartograph_graph | Callers, callees, reverse impact, paths, and symbol similarity |
| cartograph_affected | Bounded affected-test selection |
| cartograph_numerical | Generation-scoped static numerical hazards, coverage, explanations, and non-executing probe plans |
| cartograph_review | Git-ref plus staged, unstaged, and untracked evidence |
| cartograph_admin | Explicit index, sync, embedding, and maintenance jobs |
MCP profiles are coding, core, full, read-only, and review. Profiles
are immutable authorization ceilings with deterministic tool lists; a narrower
profile cannot call hidden tools. Cartograph supports modern stateless MCP
2026-07-28 plus the legacy 2024-11-05 initialize path. Modern hosts should
cache the stable complete catalog and dynamically place only task-relevant
schemas in the model context; the server does not mutate tools per connection.
See MCP usage.
These are selected high-use forms, not the complete command inventory. Run
cartograph --help for the installed binary or see the
CLI reference for every top-level command.
cartograph index [PROJECT]
cartograph status [PROJECT]
cartograph find <QUERY> --by auto|name|content|env|sql|build|path|refere
Truncated for display — read the full file on GitHub.
Related Skills
Agent-Reach
107.1kGive your AI agent eyes to see the entire internet. Read & search Twitter, Reddit, YouTube, GitHub, Bilibili, XiaoHongShu — one CLI, zero API fees.
claude-mem
107.0kPersistent Context Across Sessions for Every Agent – Captures everything your agent does during sessions, compresses it with AI, and injects relevant context back into future sessions. Works with Claude Code, OpenClaw, Codex, Gemini, Hermes, Copilot, OpenCode + More
headroom
78.4kCompress tool outputs, logs, files, and RAG chunks before they reach the LLM. 20% fewer tokens for coding agents, 60-95% fewer tokens for JSON, same answers. Library, proxy, MCP server.
CowAgent
47.7kOpen-source personal AI assistant & Agent Harness. Plans tasks, runs tools and skills, self-evolves with memory and knowledge. Multi-agent, multi-model, multi-channel. Lightweight, extensible, one-line install.
Languages
Trust signals
From repository metadata: license, adoption, age and documentation. Not a code audit — see the Safety scan above for what the skill file itself contains.
