Stats
Actions
Tags
From ac-workflow
Validates SPEC against PLAN.md instructions, evaluates if # AI Section > ## Plan achieves objectives aligned with research, amends for compliance, summarizes changes, and commits via script.
How this agent operates — its isolation, permissions, and tool access model
Agent reference
ac-workflow:agents/spec/validate-inlineThe summary Claude sees when deciding whether to delegate to this agent
VALIDATE SPEC PLAN 1. COMPLIES with agents/spec/PLAN.md EXACT instructions. 2. Will successfully ACHIEVE Objectives and Details, ALIGNED with Research. SPEC: $ARGUMENT - COMMIT ONLY the files you changed. 1. RE-READ SPEC. 2. EVALUATE if SPEC COMPLIES with agents/spec/PLAN.md EXACT instructions. 3. EVALUATE if AFTER implementing `# AI Section > ## Plan`, the SPEC implementation will successfully...
VALIDATE SPEC PLAN
SPEC: $ARGUMENT
# AI Section > ## Plan, the SPEC implementation will successfully ACHIEVE Objectives and Details, ALIGNED with Research.# AI Section > ## Plan AGAINST# AI Section > ## Plan that does not align with ## Tasks (L1) requirement## Tasks (L1)# Source spec resolver (plugin-aware)
source "${CLAUDE_PLUGIN_ROOT}/scripts/spec-resolver.sh"
# Commit spec changes
commit_spec_changes "<spec_path>" "VALIDATE_INLINE" "<NNN>" "<title>"
CLAUDE.md (AGENTS.md).npx claudepluginhub waterplanai/agentic-config --plugin ac-workflowValidates spec plans to confirm they achieve objectives and align with research. Evaluates AI plan against full spec, researches files, appends formatted feedback, summarizes, and commits spec changes.
Verifies development plans achieve goals before execution using goal-backward analysis. Checks requirement coverage, task completeness, dependencies, scope sanity, and SPEC compliance. Read-only.
Structurally validates implementation plans before execution across 7 dimensions: task completeness, verify quality, dependencies, scope sanity, read-first validity, spec alignment, test coverage.