post-product-questions
Skill2 starsChanged 4 months agoMITView on GitHub
Phase 1 engg-reviews (PE): after /prd-codebase-map, format product questions (scenario, example, recommendation, why, alternatives) and post them on the open Meta PRD PR with an explicit request for PM feedback. Does not decide product answers, does not edit the PRD, does not run PM skills. Gate- independent — never sets impact-map-* labels. PM later pulls the thread, updates PRD/outline, and re-runs validate-requirements / update-documents.
In this repository22 other files it gives its agents.
- ground-specSkill
- initiative-feasibilitySkill
- learning-extractSkill
- loop-specSkill
- pre-implementSkill
- purge-initiative-artifacts-appSkill
- spec-draftSkill
- spec-implementation-planSkill
- spec-technical-reviewSkill
- ensure-repo-graphSkill
- prd-codebase-mapSkill
- review-product-questionsSkill
- commit-workspaceSkill
- create-board-ticketsSkill
- open-draft-prSkill
- prd-impact-mapSkill
- prd-qualitySkill
- prd-thinkSkill
- purge-initiative-artifacts-metaSkill
- review-findingsSkill
- update-documentsSkill
- validate-requirementsSkill
The file
We split this skill into the rules it gives an agent
6 rules. About 0 are common, 0 are distinctive, and 6 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈0
Distinctive · estimate
This author's own choices. Worth a look.
6
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 24
This is not a PM decision skill: Do not walk AskQuestion accept/reject loops.
- Very specific: line 47
Never set/clear impact-map-* or spec-* labels.
- Very specific: line 66
If stance file exists, overlay pe_action / refined recommendations.
- Very specific: line 87
If gh unavailable: print the file path and exact command.
- Very specific: line 111
Use contract: engg-reviews/v1 and stage post-product-questions.
- Very specific: line 113
Always gate_coupled: false.
≈ “Common” is an estimate: at least 1 in 5 of the 372,429 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Post product questions (PE → Meta PR)
PE closes the engg-reviews loop by publishing codebase-grounded product questions on the Meta PRD PR and asking PM for feedback.
This is not a PM decision skill. Do not walk AskQuestion accept/reject
loops. Do not edit prd/INIT-*.md or outlines. Do not run
/validate-requirements, /review-findings, or /update-documents.
Question shape: ../references/product-question-template.md.
Handoff: ../references/handoff-adjunct.md.
Lane split
| Role | Action |
| PE | Map → optional /review-product-questions → this skill posts on Meta PR |
| PM (later, meta) | Read PR comments → update PRD + outline → requirements skills |
| Gate 1 | Unchanged — still impact-map-lgtm + Approve; this post does not unlock it |
NON-NEGOTIABLE
- Input is one
/prd-codebase-mapartifact — do not re-run graphs. - Include for each Q: scenario, example, recommendation, why, alternatives, evidence (so PM can decide without reading code).
- No product decisions in this skill — recommendations are PE proposals only.
- No PRD / outline edits.
- GitHub comment (or draft body file) only after explicit PE authorization.
- Never set/clear
impact-map-*orspec-*labels. gate_coupled: falsealways.- Dual output: local body file under pe-workspace
out/reports/+ Meta PR comment when authorized.
Inputs
- Map path — (REQUIRED)
PRD-Codebase-Map-{INIT}.md - PE stance — (OPTIONAL)
PE-Product-Stance-{INIT}.mdfrom/review-product-questions. When present, prefer refined recommendations and omit dropped questions. - Meta PR — (REQUIRED) number/URL + repo (e.g.
autrio10x/drivestream-meta#104) - Mode — all questions / conflicts-only / PE-selected IDs
Process
T0 — Gather
- Parse § Product questions from the map.
- If stance file exists, overlay pe_action / refined recommendations; skip
drop-questionrows. - Resolve Meta PR via
gh pr viewwhen available. - Cap: post the same capped set as the map (default ≤10).
T1 — Format PE feedback request
Write out/reports/Meta-PR-Product-Questions-{INIT}.md using
references/output-template.md.
Tone: PE proposing grounded questions + recommended stance; asking PM to confirm, adjust, or reject and then update PRD/outline.
T2 — Authorize + post
Show the body. Ask PE to authorize Meta PR comment.
gh pr comment {N} --repo {org}/{meta} --body-file out/reports/Meta-PR-Product-Questions-{INIT}.mdIf gh unavailable: print the file path and exact command; outcome blocked
until posted or PE marks posted-manually.
T3 — Handoff
| Outcome | When | next_candidates |
pass | Comment posted (or PE confirms manual post) | [] — wait for PM |
needs-input | Meta PR / map missing | [] |
blocked | No auth / gh failure | post-product-questions |
skipped | PE declines to post | [] |
After this skill (PM lane — not engg-reviews)
- PM reads Meta PR thread.
- PM updates PRD + outline.
- PM re-runs requirements skills (
validate-requirements, etc.). - PE may later re-run
/prd-codebase-mapon a revised PRD (new map revision). - Gate 1 remains independent.
Workflow handoff
Append the envelope from
../references/handoff-adjunct.md.
Use contract: engg-reviews/v1 and stage post-product-questions.
Always gate_coupled: false. external_action: true only when a Meta PR
comment is the authorized next step (comment ≠ gate labels).
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

