agentleFS
hussainatyhussainaty/ai-agent-skills/skills/remote-agent-graphify/SKILL.md

remote-agent-graphify

Skill1 starsChanged 31 days agoMITView on GitHub

Prepare and maintain Graphify repository context for coding agents running in Coder workspaces or the TrueForge harness. Use for repository exploration, architecture questions, and context refreshes; not for provisioning either platform.

The file

We split this skill into the rules it gives an agent

9 rules. About 0 are common, 7 are distinctive, and 2 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈7

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 2

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Keep a durable, queryable Graphify map beside the repository while Coder supplies the development workspace and TrueForge supplies an …

    line 8
  2. Very specific:

    Coder workspace: use the checked-out workspace directory.

    line 15
  3. Distinctive:

    Do not provision a Coder workspace, install a dependency, publish a TrueForge skill, connect an MCP server, or change sandbox permissions …

    line 27
  4. Very specific:

    If graphify-out/graph.json is absent, use the installed Graphify workflow to build a baseline graph for that root.

    line 38
  5. Distinctive:

    If a graph already exists and code or docs changed, run Graphify's incremental update flow.

    line 41
  6. Distinctive:

    If the user asks an architecture question and a graph already exists, query it immediately.

    line 44
  7. Distinctive:

    When Graphify is not available in the selected runtime, state that the missing dependency is the Graphify skill, CLI, or MCP integration.

    line 48
  8. Distinctive:

    Use the Graphify workflow for repository structure and evidence-backed codebase retrieval.

    line 74
  9. Distinctive:

    Use the same repository revision when passing work from a Coder workspace to a TrueForge sandbox.

    line 84

≈ “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

Remote Agent Graphify

Keep a durable, queryable Graphify map beside the repository while Coder supplies the development workspace and TrueForge supplies an optional agent runtime. The graph is repository knowledge; it is not a substitute for either platform's skill, sandbox, approval, or session mechanisms.

Select the runtime

  • Coder workspace: use the checked-out workspace directory. This skill lives in .agents/skills/, which Coder discovers when that directory is the workspace working directory. After changing a skill, refresh the Coder chat's context or start a new chat before expecting the change to apply.
  • TrueForge: mount this directory from a Git repository as a Git-backed skill. Ensure the selected sandbox has access to the target repository and a shell or an approved Graphify MCP tool. Treat generated graph data as ephemeral unless the sandbox or checkout explicitly persists it.

For each platform's skill discovery paths, size limits, and mounting details, read references/platform-handoff.md.

Do not provision a Coder workspace, install a dependency, publish a TrueForge skill, connect an MCP server, or change sandbox permissions without the user's authorization for that platform action.

Keep one graph per repository

Store generated data at <repo>/graphify-out/; never put it inside a skill directory or load graph.json wholesale into agent instructions.

  1. Resolve the repository root. Prefer the Git top-level directory; otherwise use the workspace project root.
  2. If graphify-out/graph.json is absent, use the installed Graphify workflow to build a baseline graph for that root. Report the corpus summary, graph health, and Graphify's output paths.
  3. If a graph already exists and code or docs changed, run Graphify's incremental update flow. Preserve the manifest and cache so unchanged files are not re-extracted.
  4. If the user asks an architecture question and a graph already exists, query it immediately. Do not rebuild first. Expand the query against the graph vocabulary and cite source locations returned by Graphify.

When Graphify is not available in the selected runtime, state that the missing dependency is the Graphify skill, CLI, or MCP integration. Do not invent graph findings from an absent graph or silently replace Graphify with an unrelated indexer.

Assemble agent context deliberately

For a focused implementation task, begin with the named files or symbols and select one or two graph hops, with a 2,500-token budget by default. For an architecture, roadmap, or cross-repository task, begin with community summaries and bridge nodes, using a 4,000-token budget by default. Keep extracted facts, inferences, and proposed work distinct.

Give the agent only the selected graph context and the source snippets needed to verify it. Keep the graph output on the workspace volume or persistent checkout so later Coder or TrueForge runs can refresh and query the same repository map.

Hold the selected context in a working notebook (the working-context-notebook skill) on the same workspace volume or persistent checkout. Record the repository commit, graph build time, and the selected nodes as sourced facts. A later run in either platform then reads the accepted revision back and can check whether the graph is stale before trusting it.

Use the companion workflows

Use the Graphify workflow for repository structure and evidence-backed codebase retrieval. For time-sensitive public information that changes an engineering decision, use scrapegraphai-research first and retain only verified primary sources as graph evidence. For a user-authorized implementation, use product-quality-loop to verify the real user journey after the graph-guided change. Use agent-workflow-pipeline when the task needs all of these routes plus optional Jev or Needle 3 model routing.

Handoff between Coder and TrueForge

Use the same repository revision when passing work from a Coder workspace to a TrueForge sandbox. Before acting on a prior graph, confirm that its manifest or last update matches the checkout; otherwise update it. Keep the agent's model, tools, sandbox, approvals, and session state in TrueForge, while Graphify provides repository structure and evidence-backed retrieval.

For proposed engineering work, write graphify-out/context.json and graphify-out/CONTEXT_PLAN.md only when the user requests planning or architecture context. Record the token budget, source evidence, uncertainties, risks, extension points, and validation signals there.

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.