A1 checklist

Skill N3URAL-A1/a1-specforge/skills/a1-checklist

Use PROACTIVELY as the pre-flight readiness gate before ANY feature implementation starts. Runs 10 deterministic structural checks on a feature's wave-plan: spec status is `clarified`, wave-plan exists, every wave has a Suggested agent line, dependencies form a DAG (no cycles), every wave references advanced stories, project CLAUDE.md is present, plans/ directory convention is honored, plan frontmatter has all required fields, FR coverage between spec and plan is bijective (check #9: every spec FR in exactly one wave, no phantom FRs), and the plan's spec_path links back to the spec (check #10). Checks #9/#10 are the spec↔plan consistency gate (the former a1-check skill, retired M13) — a1-new-feature's Phase 4.5 runs exactly this subset via `--only 9,10` (exit 0/1/2). Severity maps to BLOCKER (stops the gate), MAJOR (warns), MINOR (info). MUST trigger on: "checklist for <feature>" (alias: "checkliste für <feature>"), "pre-flight check", "is the plan implementation-ready" (alias: "ist der plan implementierungs-bereit"), "ready for implementation" (alias: "ready für implementation"), "plan check", "consistency check for <feature>" (alias: "konsistenz-check für <feature>"), "fr-coverage check", "can I start implementation now" (alias: "kann ich jetzt mit der umsetzung starten"), "check wave-plan completeness" (alias: "prüf die wave-plan-vollständigkeit"), "checklist run <slug>", "a1-checklist", or any request to verify whether a wave-plan is ready to be handed to executing agents. Exit semantics: 0=PASS or PASS_WITH_WARNINGS, 1=FAIL (BLOCKER), 2=ERROR (setup). Do NOT activate for: semantic spec review (→ a1-rene-requirement-engineer), generic project audits (→ a1-analyze), or post-implementation quality checks (→ a1-tobi-tester).From its SKILL.md

Install

npx -y skills add N3URAL-A1/a1-specforge --skill a1-checklist

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

2 stars2 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

SKILL.md

3.6 KB, as published. Nobody here has run it

a1-checklist — Pre-Flight Readiness Gate

Language: English-first; German trigger aliases supported.

Thin Markdown wrapper around the deterministic CLI gate. All logic lives in <repo>/_shared/a1-tools.cjs checklist run. The workflow file translates the structured result into user-facing guidance (user's language — see _shared/language-policy.md).

When to use

Activate when the user wants to know whether a feature's wave-plan is implementation-ready. The check runs across:

The feature's spec (projects/<slug>/spec/<###>-<feature-slug>.md)
The feature's wave-plan (projects/<slug>/plans/<###>-<feature-slug>-wave-plan.md)
The project root metadata (projects/<slug>/CLAUDE.md, plans/ layout)

When NOT to use

Semantic review of a spec's content → a1-rene-requirement-engineer
Cross-cutting project audit → a1-analyze
After implementation, retrospective quality check → a1-tobi-tester

Hand-off

PASS → user proceeds to implementation
PASS_WITH_WARNINGS → user can proceed but warnings are listed for awareness
FAIL (BLOCKER) → workflow suggests fix-paths (re-clarify spec, re-plan, or fix the dep cycle), then stops
ERROR → workflow asks user to fix setup (missing files / bad frontmatter)

Hard rules

This skill is read-only. It never edits spec, plan, or CLAUDE.md files.
The skill never gates on semantic content — only structural and metadata invariants. Story-level review belongs to humans (or Rene).
All check logic is in _shared/a1-tools.cjs checklist run — keep this skill thin. If you find yourself reimplementing a check here, move it to the CLI.

Workflows

workflows/01-run.md — Run the checklist and interpret the result.