agentleFS
attilaszaszattilaszasz/sdd-pilot/.agents/skills/sddp-autopilot/SKILL.md

sddp-autopilot

Skill98 starsChanged 28 days agoMITView on GitHub

Run the full SDD feature-delivery pipeline. Direct command-bar dispatch only; do not select for general queries.

The file

We split this skill into the rules it gives an agent

3 rules. About 0 are common, 2 are distinctive, and 1 is very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈2

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 1

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Never simulate implementation, QC, test results, or marker creation.

    line 13
  2. Very specific:

    After Clarify or its skip path, the canonical workflow creates a separate ephemeral P1_REQUIREMENT_SNAPSHOT from the live spec.md.

    line 17
  3. Distinctive:

    When any sub-skill says Delegate, read the exact referenced sub-agent file at that point, not before, then perform the delegated task …

    line 28

≈ “Common” is an estimate: at least 1 in 5 of the 372,429 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

Argument hint: [optional: feature description; omit to select the first unchecked epic] Command category: orchestration Prerequisites: autopilot:enabled, product-document:planning-ready, technical-context:planning-ready

You are running the Autopilot Pipeline — a fully automated SDD workflow that executes all phases (Specify → Clarify → Plan → Checklist → Tasks → Analyze → Implement+QC) in a single uninterrupted turn without user interaction. Every decision point, phase lifecycle event (start, complete, skip), gate check, and halt is logged to autopilot-log.md using a structured 7-column schema (Timestamp | Phase | Event | Detail | Outcome | Rationale | Artifacts). Every artifact or document mentioned in a log row must appear as a clickable relative Markdown link in the Artifacts column. At run end, a ## Run Summary section is appended with per-phase status and links to final artifacts.

Autopilot is real unattended execution, not a demo, showcase, dry run, or simulation. Execute each phase for real: perform actual file edits, actual build/test/lint/QC commands, and create artifacts only when the owning phase has genuinely completed. Never simulate implementation, QC, test results, or marker creation. If real execution cannot complete in the current environment, halt and report the blocker.

Load and follow the workflow in .github/sddp/workflows/autopilot-pipeline/WORKFLOW.md. Retain the initial full Context Gatherer report as PIPELINE_CONTEXT and pass it unchanged to every inline phase; downstream phases re-check mutable artifacts instead of delegating Context Gatherer again. After Clarify or its skip path, the canonical workflow creates a separate ephemeral P1_REQUIREMENT_SNAPSHOT from the live spec.md; it is not part of PIPELINE_CONTEXT and is passed only to Tasks and fresh Implement+QC gates after checksum verification.

The pipeline skill will instruct you to load and execute these sub-skills inline, in order:

  1. Specify → .github/sddp/workflows/specify-feature/WORKFLOW.md
  2. Clarify → .github/sddp/workflows/clarify-spec/WORKFLOW.md
  3. Plan → .github/sddp/workflows/plan-feature/WORKFLOW.md
  4. Checklist → .github/sddp/workflows/generate-checklist/WORKFLOW.md (looped until queue exhausted)
  5. Tasks → .github/sddp/workflows/generate-tasks/WORKFLOW.md
  6. Analyze → .github/sddp/workflows/analyze-compliance/WORKFLOW.md
  7. Implement+QC → .github/sddp/workflows/implement-qc-loop/WORKFLOW.md

When any sub-skill says Delegate, read the exact referenced sub-agent file at that point, not before, then perform the delegated task yourself.

AUTOPILOT = true for all phases. At every user interaction point, choose the recommended default and log the decision — never prompt the user.

Report progress at each phase boundary. Only halt for the conditions defined in the pipeline skill.

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.