agentleFS
osocietyosociety/vernet/.github/skills/project-task/SKILL.md

project-task

Skill566 starsChanged 13 days agoApache-2.0View on GitHub

Complete Vernet feature work, bug fixes, release tasks, workflow issues, and repo-wide implementation work by using the project’s instructions, memory, and the most relevant specialized skill.

The file

We split this skill into the rules it gives an agent

12 rules. About 0 are common, 7 are distinctive, and 5 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.

  • 5

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Use this when a request is broad or cross-cutting and needs the repo’s guidance, memory, and the most relevant specialized skill to …

    line 9
  2. Very specific:

    Read the repository guidance first: .github/copilot-instructions.md, AGENTS.md, and ARCHITECTURE.md before making a structural or …

    line 13
  3. Distinctive:

    Check relevant memory before deciding the approach:

    line 14
  4. Very specific:

    read /memories/ for user or repo constraints, testing patterns, and past fixes

    line 15
  5. Very specific:

    use /memories/session/ for multi-step task tracking

    line 16
  6. Distinctive:

    Keep changes narrow and in the correct layer:

    line 28
  7. Very specific:

    prefer flutter test path/to/test.dart for targeted checks

    line 36
  8. Very specific:

    run flutter analyze after coherent Dart changes

    line 37
  9. Distinctive:

    Treat the app as a Flutter network-diagnostics project with a layered architecture.

    line 43
  10. Distinctive:

    When a task touches project memory or historical debugging patterns, incorporate the relevant stored guidance as part of the working plan.

    line 44
  11. Distinctive:

    Do not claim a check passed without fresh output.

    line 45
  12. Distinctive:

    If local verification is blocked by environment limits, report the exact blocker and what remains unverified.

    line 46

≈ “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

Project Task

Use this when a request is broad or cross-cutting and needs the repo’s guidance, memory, and the most relevant specialized skill to complete the work without guessing.

Procedure

  1. Read the repository guidance first: .github/copilot-instructions.md, AGENTS.md, and ARCHITECTURE.md before making a structural or architectural change.
  2. Check relevant memory before deciding the approach:
    • read /memories/ for user or repo constraints, testing patterns, and past fixes
    • use /memories/session/ for multi-step task tracking
    • preserve prior notes about failing-test workflow and verification standards
  3. Identify the task type and choose the closest specialized skill:
    • graphify for architecture, file relationships, request routing, or “where is this implemented?” questions
    • github-issue-implementation for issue-driven fixes or requested behavior changes
    • flutter-feature-iteration for UI, screens, and feature behavior changes
    • flutter-test-repair for failing unit, widget, or integration tests
    • github-actions-repository-workflows for CI workflow or repo automation failures
    • android-fastlane-release for Android release or beta delivery work
    • firebase-hosting-config for Firebase or web deployment config problems
    • close-chat when asked to close or wrap up a chat, record durable context, or prepare a handoff
  4. State one falsifiable cause or expected-behavior hypothesis, then pick the smallest validation that could disprove it. Start from the concrete problem, not a broad refactor.
  5. Keep changes narrow and in the correct layer:
    • UI in lib/pages/ or lib/widgets/
    • networking in lib/services/
    • data in lib/models/
    • utilities in lib/utils/
    • feature state in the local or existing project pattern
  6. Preserve user work and unrelated dirty files. Never discard or overwrite changes without explicit reason.
  7. Validate with the smallest relevant command and rerun after the first meaningful edit:
    • prefer flutter test path/to/test.dart for targeted checks
    • run flutter analyze after coherent Dart changes
    • for CI or platform work, use the corresponding repository command rather than a broad suite
  8. Report the outcome with a short summary of the task, root cause or behavior implemented, files changed, validation commands and results, and any remaining limitations.

Vernet-specific expectations

  • Treat the app as a Flutter network-diagnostics project with a layered architecture; do not mix UI and service logic.
  • When a task touches project memory or historical debugging patterns, incorporate the relevant stored guidance as part of the working plan.
  • Do not claim a check passed without fresh output.
  • If local verification is blocked by environment limits, report the exact blocker and what remains unverified.

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.