agentleFS
selamy-labsselamy-labs/agent-skills/skills/codegraph-bootstrap/SKILL.md

codegraph-bootstrap

Skill0 starsChanged 3 days agoMITView on GitHub

Use when starting code exploration or implementation in a repository where CodeGraph may be available. Ensure CodeGraph is installed/initialized when authorized, then prefer CodeGraph queries for semantic code context.

  • Pipes a download into a shell

The file

Show its rules and text

We split this skill into the rules it gives an agent

13 rules. About 0 are common, 8 are distinctive, and 5 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈8

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 5

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Use CodeGraph as the first semantic code-navigation path when it is available.

    line 8
  2. Distinctive:

    Check for an initialized project index:

    line 14
  3. Very specific:

    If codegraph exists but the project is not initialized, run:

    line 20
  4. Distinctive:

    Do this only inside the target repository.

    line 26
  5. Very specific:

    If codegraph is missing, look for the environment's declared installer or package source before using an ad hoc install command.

    line 30
  6. Distinctive:

    Not authorized: a production/shared host with no declared installer, a read-only review task, an environment where dependency installation …

    line 44
  7. Distinctive:

    When authorized, prefer the environment's declared installer.

    line 48
  8. Distinctive:

    After installation or initialization, prefer:

    line 56
  9. Very specific:

    If the MCP tool is available, use codegraph_explore instead of the CLI for the same semantic lookup.

    line 64
  10. Distinctive:

    Keep installation declarative where the environment has a dotfiles, Nix, package-manager, image-build, or IaC source of truth.

    line 69
  11. Very specific:

    Do not run codegraph install --yes on a host whose agent config is source-controlled unless that mutation is also codified or the …

    line 72
  12. Distinctive:

    Do not block the task if CodeGraph cannot be installed or initialized.

    line 75
  13. Very specific:

    Treat .codegraph/ as local generated index state.

    line 77

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

CodeGraph Bootstrap

Use CodeGraph as the first semantic code-navigation path when it is available. The goal is to give the agent indexed symbol, call-path, and impact context before falling back to file-by-file grep/read exploration.

Workflow

  1. Check for an initialized project index:

    codegraph status .
  2. If codegraph exists but the project is not initialized, run:

    codegraph init .

    Do this only inside the target repository. Do not initialize a home directory, filesystem root, vendored dependency cache, or generated build directory unless the user explicitly asks.

  3. If codegraph is missing, look for the environment's declared installer or package source before using an ad hoc install command. Check repo docs, bootstrap scripts, dotfiles, Nix flakes, package manifests, container images, or other source-owned provisioning files that already mention CodeGraph.

    rg -n "codegraph|CODEGRAPH_" . "$HOME/.config" 2>/dev/null
  4. Install CodeGraph only when authorized:

    • Authorized: the user asked for installation, the task is explicitly a bootstrap/provisioning task, or a source-owned installer exists for the environment.
    • Not authorized: a production/shared host with no declared installer, a read-only review task, an environment where dependency installation is disallowed, or any case where the user asked not to change the host.
  5. When authorized, prefer the environment's declared installer. If no declared installer exists but the user explicitly authorizes a direct user-local install, use the upstream installer:

    curl -fsSL https://raw.githubusercontent.com/colbymchenry/codegraph/main/install.sh | sh
  6. After installation or initialization, prefer:

    codegraph explore "question or symbol area" --path .
    codegraph node SymbolName --path .
    codegraph impact SymbolName --path .

    If the MCP tool is available, use codegraph_explore instead of the CLI for the same semantic lookup.

Guardrails

  • Keep installation declarative where the environment has a dotfiles, Nix, package-manager, image-build, or IaC source of truth. Add or update that source instead of leaving a live-only mutation.
  • Do not run codegraph install --yes on a host whose agent config is source-controlled unless that mutation is also codified or the host-local file is intentionally outside shared config.
  • Do not block the task if CodeGraph cannot be installed or initialized. State the reason, then continue with normal repository tools.
  • Treat .codegraph/ as local generated index state. Do not commit it unless the repository explicitly tracks it.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.