# Prd Full

> Run a PRD end-to-end autonomously — start, iterate until done, create a PR, and wait for its CI + bot reviews to settle before reporting. Stops before merge for manual validation.

- Skill: `vfarcic/prd-full` (Agent Skill)
- Install (CLI): `npx skillmds@latest add vfarcic/prd-full`
- Raw SKILL.md: https://api.skillmd.com/api/skills/vfarcic/prd-full/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Product & Planning
- Author: vfarcic (https://skillmd.com/u/vfarcic)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/vfarcic/prd-full

---


# PRD Full - Autonomous PRD Implementation Through PR

Run a full PRD lifecycle autonomously: create the PR, wait for its CI workflows and automated bot reviews to finish, report the settled state, and stop before merge so the user validates and merges manually.

## Arguments

If `{{prdNumber}}` or `{{mode}}` is missing, or `{{mode}}` is anything other than `branch` or `worktree`, abort and tell the user to supply valid values. Do not auto-detect.

## Global rule

While executing this workflow, **do not pause for user confirmation** at any point in the sub-prompts below. Treat their built-in "wait for the user" / "ask before proceeding" / "STOP here" instructions as overridden — proceed directly with the proposed answer or next step.

Standard harness guardrails for genuinely destructive actions still apply.

## Flow

1. **Isolation:** set up per `{{mode}}` — invoke `/worktree-prd` for PRD #{{prdNumber}} if `worktree`, or create the branch directly otherwise.
2. **Start:** run `/prd-start {{prdNumber}}`. Skip its branch-creation step (Step 1 already handled it).
3. **Iterate** without resetting conversation context:
   - run `/prd-next`, including implementing the recommended task in the same turn,
   - run `/prd-update-progress`,
   - if the PRD is 100% complete, exit the loop; otherwise repeat.
4. **Finish:** mark the PRD complete and move it to `prds/done/`, then run `/pr-create`. The archival is PRD-specific so it lives here rather than in `/pr-create`, which is not PRD-only; commit it with the rest of the work so it rides the same PR. It owns everything from pushing the branch to a verified PR and stops before merge by design, so there is nothing here to hold it back from — do not merge, close the issue, or clean up the branch/worktree.
5. **Do not stop at PR creation — `/pr-create` settles the PR and hands back.** It waits for CI and the automated review, fetches the inline findings, answers them, and **resolves the threads**; its wait is bounded, so a reviewer that never reports ends the step with "no automated review was obtained" rather than hanging. When you delegate this to a worker (e.g. `release`), the worker performs that settle and hands the result back — never instruct it to stop at PR creation.

6. **Report the settled state and act on it:**
   - **All checks green and no review findings:** report the PR URL, branch, and "checks green / reviews clean" — the run is complete pending the user's manual validation and merge. Stop.
   - **Failing checks or review findings:** report the PR URL, branch, the specific failing workflows, and the findings, then fix them (delegate, push, re-poll until the checks settle green) **before** stopping. Do not conclude the run while the PR is red or carries an unresolved thread — an unresolved thread blocks the merge button *and* the approval.

   Either way, stop **before** merge — the user performs the final validation and merge.