Skip to main content

sdd-4a-validate

Validate plan readiness before implementation with parallel quality gates. Use when: a plan has been created by the Architect phase and you want to verify completeness, doctrine compliance, and dependency ordering before starting implementation.

Quellinformationen

Repository
willvelida/biotrackr
Letzte Quellaktivität
22. August 2026 um 06:02
Erkannte Sprache von SKILL.md
Englisch
Sterne
6
Forks
3

Installationsoptionen

Standardmäßig ist der Prompt ausgewählt, der zuerst die Quelle prüft. Sie können zu einem direkten Befehl wechseln oder eine lokale Kopie herunterladen.

Quelldateien prüfen

Lesen Sie SKILL.md und alle von SkillsMP angezeigten Begleitdateien, bevor Sie sich für eine Installation entscheiden.

SKILL.md wird angezeigt

SKILL.md
Quellanweisungen · Schreibgeschützte Vorschau
name
sdd-4a-validate
description
Validate plan readiness before implementation with parallel quality gates. Use when: a plan has been created by the Architect phase and you want to verify completeness, doctrine compliance, and dependency ordering before starting implementation.
argument-hint
[slug=...] [plan=...]
# SDD Phase 4a: Validate Plan > [!CAUTION] > Do NOT modify any source files during this phase. This is a READ-ONLY validation command. Validate that a plan is ready for implementation by running parallel quality gates. ## Inputs * **slug** (Optional): Slug from prior phases. Inferred from context if omitted. * **plan** (Optional): Path to the plan file from Phase 4 (Architect). ## When to Read References Load only the reference the current problem needs: * Task table format the plan must conform to: `.github/instructions/sdd-conventions.instructions.md` * Whether the complexity score is defensible: `docs/standards/harness-governance.md` * Whether the plan cites real build and test commands: `AGENTS.md` ## Step 0: Doctrine Resolution Before starting, resolve project conventions: 1. Check for project doctrine files: * `docs/project-rules/` (constitution.md, rules.md, idioms.md, architecture.md) * `copilot-instructions.md`, `AGENTS.md`, `CONTRIBUTING.md`, `README.md` 2. If no doctrine found, scan the codebase for dependency manifests, build systems, and directory patterns. 3. Extract: build command, test command, coverage threshold, naming conventions, CI platform. 4. Unknown values become explicit `[TODO]` markers — never assume silently. ## Step 1: Load Plan Context 1. Read the plan file from `{plan}` or `.copilot-tracking/plans/{date}/{slug}/{slug}-plan.md`. 2. Read the specification for cross-reference. 3. Read the research dossier if it exists. 4. Read any workshop documents or ADRs created during this cycle. ## Step 2: Launch Parallel Validators Run these validators concurrently: 1. **Structure Validator** — verify the plan has all required sections: Purpose, Complexity Score, Goals, Non-Goals, Acceptance Criteria, Progress, Phases with task tables, Validation. 2. **Completeness Validator** — verify every acceptance criterion from the spec maps to at least one task in the plan. Flag unmapped criteria. 3. **Doctrine Validator** — verify the plan respects doctrine boundaries (naming conventions, module topology, service lifetimes, API patterns). 4. **Dependency Validator** — verify task dependencies are acyclic and phase ordering is logical. Flag tasks that reference files in later phases. Each validator produces findings with severity (CRITICAL, HIGH, MEDIUM, LOW). ## Step 3: Synthesize Verdict Aggregate all findings and issue a verdict: * **READY** — zero CRITICAL or HIGH findings. Proceed to implementation. * **NOT READY** — one or more CRITICAL or HIGH findings. List specific fixes needed. Present the findings table: ```markdown ## Validation Report ### Verdict: {READY | NOT READY} | # | Severity | Validator | Finding | Fix Required | |---|----------|-----------|---------|--------------| | 1 | {severity} | {validator name} | {description} | {specific fix} | ``` ## Step 4: Present Verdict Present the verdict and findings to the user. If NOT READY, the user can: 1. Fix the issues and re-run validation. 2. Override with explicit acknowledgment (proceed despite issues). --- > [!IMPORTANT] > **STOP.** Present the validation verdict. Do not proceed to implementation. The user decides whether to fix issues, override, or proceed.
Auf GitHub ansehen