agentleFS
hushh-labshushh-labs/hushh-research/.agents/skills/repo-context/SKILL.md

repo-context

Skill40 starsChanged 4 days agoApache-2.0View on GitHub

Use when a request starts with scanning the Hussh repository, establishing repo context, mapping code/docs/skills, or deciding which owner skill should own the next step.

4 other projects use a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

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

  • 0

    Common

    Most of the 4 edited versions we read keep these.

  • 0

    Distinctive

    This author's own choices. Worth a look.

  • 9

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Avoid overlap with frontend, backend, security-audit, and docs-governance.

    line 12
  2. Very specific:

    Run summary first to get owners, spokes, surface coverage, and uncovered surfaces.

    line 56
  3. Very specific:

    Use list-workflows and route-task <workflow-id> when the request matches a recurring execution shape.

    line 58
  4. Very specific:

    Use data-model-audit before data-model implementation when migration/table classification, retention posture, or legacy write drift is in …

    line 59
  5. Very specific:

    Use impact <workflow-id> before implementation when blast radius, docs, or verification scope is unclear.

    line 60
  6. Very specific:

    Keep the first pass compact by default.

    line 61
  7. Very specific:

    Use --text when a human-readable low-token routing summary is enough.

    line 62
  8. Very specific:

    Use validate or audit when changing the taxonomy, coverage map, workflow packs, or skill-routing docs.

    line 64
  9. Very specific:

    Treat low-severity context-size and modularity findings from audit as review-required signals only.

    line 65

“Common” is measured: kept by at least 6 in 10 of the 4 edited versions of this file we could read. “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

Hussh Repo Context Skill

Purpose and Trigger

  • Primary scope: repo-context-intake
  • Trigger on broad repository scans, cross-domain mapping, surface discovery, and choosing the correct owner skill before implementation work starts.
  • Avoid overlap with frontend, backend, security-audit, and docs-governance.

Coverage and Ownership

  • Role: owner
  • Owner family: repo-context

Owned repo surfaces:

  1. README.md
  2. data

Non-owned surfaces:

  1. frontend
  2. mobile-native
  3. backend
  4. security-audit
  5. docs-governance
  6. repo-operations
  7. analytics-observability-governance

Do Use

  1. First-pass repository orientation before any narrower implementation workflow.
  2. Mapping the full skill taxonomy and the meaningful repo surface coverage.
  3. Routing recurring work into explicit workflow packs and impact bundles.
  4. Finding code/docs/skill drift or orphaned areas in the repository map.
  5. Running the data-model audit workflow when table-family drift, migration classification, or runtime DB ownership is the starting question.

Do Not Use

  1. Narrow work that is already clearly inside a domain owner skill.
  2. Docs-only governance once the correct docs home is already obvious.
  3. Deep product, ops, or security implementation work.

Read First

  1. .agents/skills/repo-context/references/index-contract.md
  2. .agents/skills/repo-context/references/ownership-map.md
  3. docs/development/repository-map.md

Workflow

  1. Run summary first to get owners, spokes, surface coverage, and uncovered surfaces.
  2. Open exactly one deeper section next: docs, frontend, backend, skills, or commands.
  3. Use list-workflows and route-task <workflow-id> when the request matches a recurring execution shape.
  4. Use data-model-audit before data-model implementation when migration/table classification, retention posture, or legacy write drift is in scope.
  5. Use impact <workflow-id> before implementation when blast radius, docs, or verification scope is unclear.
  6. Keep the first pass compact by default; use --verbose only when a heavy section such as skills or backend needs deeper inspection.
  7. Use --text when a human-readable low-token routing summary is enough.
  8. Route the task into the recommended owner skill as soon as the domain is clear.
  9. Use validate or audit when changing the taxonomy, coverage map, workflow packs, or skill-routing docs.
  10. Treat low-severity context-size and modularity findings from audit as review-required signals only; do not split files or add folders unless a bounded owner, workflow, or navigation problem is proven.

Handoff Rules

  1. Route broad frontend work to frontend, native work to mobile-native, backend work to backend, trust and audit work to security-audit, docs work to docs-governance, analytics observability work to analytics-observability-governance, and generic ops work to repo-operations.
  2. Route licensing and notice-governance work to oss-license-governance.
  3. Route bootstrap, devcontainer, doctor, and contributor-first-run work to contributor-onboarding.
  4. Route optional-mirror sync and subtree-governance work to subtree-upstream-governance.
  5. Route skill-system work to codex-skill-authoring.
  6. Route board workflows to planning-board and public/community reply work to comms-community.

Required Checks

./bin/hushh agents scan summary
./bin/hushh agents list-workflows
./bin/hushh agents route-task repo-orientation
./bin/hushh agents impact repo-orientation
./bin/hushh agents data-model-audit
./bin/hushh agents audit
python3 .agents/skills/repo-context/scripts/repo_scan.py validate
python3 .agents/skills/repo-context/scripts/architecture_fitness.py --self-test && python3 .agents/skills/repo-context/scripts/architecture_fitness.py --limit 20

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.