refine-prd
Skill0 starsChanged 3 months agoMITView on GitHub
Turn a raw objective into a locked PRD via the deep interview protocol (Fable rounds + Codex adversarial pass) - use when the user says "refine the PRD", "let's PRD this", brain-dumps a new objective, or invokes /refine-prd.
The file
We split this skill into the rules it gives an agent
1 rule. About 0 are common, 0 are distinctive, and 1 is 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.
1
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 8
Follow standards/prd-interview.md top to bottom, working in this repo's prd.md:
≈ “Common” is an estimate: at least 1 in 5 of the 364,980 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
Refine PRD
Follow standards/prd-interview.md top to
bottom, working in this repo's prd.md:
- Capture the brain dump verbatim ("Raw objective" section of prd.md).
- Interview in rounds (AskUserQuestion, max 4 per round, recommendations attached), ordered by the frontier rule: each round asks only the questions whose prerequisite decisions are settled, and facts findable in the environment are looked up, never asked — only decisions go to the user. Continue until all 12 facets are resolved: users, workflows, data, integrations, AI features + eval criteria, screens, admin, scale, security/company data, access/deploy, success, out of scope. Update the draft after every round and show what changed.
- Adversarial pass: run
codex execwith the review prompt from the protocol; if the Codex CLI is not installed, hand over the prompt + PRD to run in the Codex app and wait for the output. - Resolve every Codex finding together; fold answers into the PRD.
- Summarize in plain language, get approval, mark prd.md locked with the
date, set
Stage: prd-locked, derive tasks.md, and offer/mockup.
Speak plainly throughout (AGENTS.md, "Working with the user"). No code or mockups until the PRD is locked.
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.

