# Prd Executor

> Deprecated compatibility shim for explicit prd-executor invocations. Use only when the user names prd-executor directly. Execute approved plans with native agent capabilities and use implementation-brief only to repair material planning gaps.

- Skill: `jmichaelschmidt/prd-executor` (Agent Skill, multi-file: 6 files)
- Install (CLI): `npx skillmds@latest add jmichaelschmidt/prd-executor`
- Raw SKILL.md: https://api.skillmd.com/api/skills/jmichaelschmidt/prd-executor/raw
- Safety review: pending (external: skill-scanner PASS, skillspector PASS)
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Product & Planning
- Author: jmichaelschmidt (https://skillmd.com/u/jmichaelschmidt)
- Updated: 2026-09-10
- Page: https://skillmd.com/skills/jmichaelschmidt/prd-executor

---


# PRD Executor Compatibility

Treat this name as deprecated. Tell the user once that approved plans now execute through native agent capabilities.

1. Read the complete plan and applicable repository instructions.
2. Confirm the objective, constraints, and verification criteria are executable.
3. If material decisions are missing, audit the plan with `implementation-brief` before mutating anything.
4. Execute in the current task using the runtime's native progress, worktree, approval, and verification behavior.
5. Delegate only when the user or applicable project instructions request it and the work has independent ownership.
6. Verify the completed outcome against the plan, not merely the individual edits.

Do not create an execution ledger, manifest, per-thread completion log, or model-routing table unless the user explicitly requests a durable external artifact.

Do not rewrite the approved plan as routine status tracking. Report deviations and unresolved verification directly to the user.