Agent skill

localsetup-agentic-prd-batch

Agentic PRD batch: when user says 'process PRDs' or 'run batch from PRD folder'; implement per spec; update status; write outcome; reference PRD schema + external-agent guide. Use when editing .agent/queue/**, prds/**, *.prd.md.

Stars 163
Forks 31

Install this agent skill to your Project

npx add-skill https://github.com/majiayu000/claude-skill-registry/tree/main/skills/other/other/localsetup-agentic-prd-batch

Metadata

Additional technical details for this skill

version
1.3

SKILL.md

Agentic PRD batch

When the user says process PRDs, run batch from PRD folder, process the queue, or similar, or when editing files in .agent/queue/ with intent to implement:

  1. Locate specs: Look for PRD/spec files in the configured path (e.g. .agent/queue/ or structured in/ under queue root). Exclude README, INDEX, SPEC-TEMPLATE. Filter by front matter status == ready (or in-progress if resuming). Sort by priority (high first), then filename date (oldest first). Version mismatch: If a spec has localsetup_framework_version and it differs from repo root VERSION, surface a warning per protocol; do not silently ignore.
  2. Implement per spec: Load each spec; follow Implementation steps and Acceptance criteria; satisfy Verification plan and Rollback plan. Use PRD schema and _localsetup/docs/PRD_SCHEMA_EXTERNAL_AGENT_GUIDE.md for format and outcome template.
  3. Update status: Set spec status: in-progress when starting; set status: done (or blocked) when finished. Update queue INDEX if present.
  4. Write outcome: Append outcome block per spec (e.g. ## Outcome or platform-specific section name, and/or aq_outcome YAML) with branch, commit SHA, files changed, verification, rollback command. See PRD_SCHEMA_EXTERNAL_AGENT_GUIDE for template.
  5. Clean-tree invariant: Before marking a spec done, ensure repo is clean (no modified tracked files except intended commits; untracked queue specs allowed). Commit or revert as needed.
  6. External confirmation: If spec front matter includes external_confirmation: acknowledged or impact_review: confirmed_by: external_agent, agent may skip human impact confirmation; otherwise follow guardrails (impact summary + user confirmation for big/destructive changes).

Reference

  • _localsetup/docs/AGENTIC_AGENT_Q_PATTERN.md - queue pattern (flat and structured layout).
  • _localsetup/docs/PRD_SCHEMA_EXTERNAL_AGENT_GUIDE.md - spec format, front matter, outcome template, clarification protocol.
  • _localsetup/docs/AGENTIC_AGENT_TO_AGENT_PROTOCOL.md - agent-to-agent handoff (ACTIVE); pre-ship gate before ship to peer.
  • _localsetup/tools/agentq_transport_client/ - version stamp CLI: python agentq_cli.py stamp-prd <path>.

Expand your agent's capabilities with these related and highly-rated skills.

Didn't find tool you were looking for?

Be as detailed as possible for better results