deck-to-issues
Skill0 starsChanged 5 months agoMITView on GitHub
Turn an approved factory-plan qWork deck into GitHub issues a single coding worker can finish in one session. Drafts one issue per feature slide, checks each with Jev (testable, too_big), dedupes against open issues, then creates them under the project map issue. Use after factory-plan is approved, or when Nick says "make tickets from the deck".
In this repository16 other files it gives its agents.
- 100-sourcesSkill
- brainstormingSkill
- building-eve-agentsSkill
- climatepulseSkill
- daily-interest-briefSkill
- delivery-deckSkill
- factory-planSkill
- factory-retroSkill
- feedback-addressingSkill
- generate-imageSkill
- helmSkill
- proofing-roomSkill
- ptSkill
- session-handoffSkill
- slide-presenterSkill
- video-ingestSkill
The file
We split this skill into the rules it gives an agent
2 rules. About 0 are common, 0 are distinctive, and 2 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.
2
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 23
Write each draft to drafts/<slug>.md in a scratch dir.
- Very specific: line 64
Every created issue has passed testable.
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 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
deck-to-issues: approved deck to worker-sized tickets
Input: an approved deck id (from factory-plan) and the project map issue number. Output: new issues on the map checklist, and no duplicates.
1. Draft
qw read <deck id> (add --at <pid> for one slide). For each feature slide, read its Acceptance text box (feature, acceptance, files hint, depends on) and draft:
Title: <feature>
Objective: <one paragraph>
Acceptance: <commands, tests or observable behaviour a gate can verify>
Files hint: <paths>
Depends on: <#issue or none>
Deck: <deck id>, slide <n>Write each draft to drafts/<slug>.md in a scratch dir.
2. Check with Jev
For each draft:
helm jev check --preset issue --file drafts/<slug>.md --jsontestableflagged (below 0.5): rewrite the acceptance into concrete checks and re-run.too_bigflagged (0.5 or above): split into two drafts along the most natural seam and re-run each.- At most 2 rewrites per draft. Still flagged: ask Nick with the draft and the flag.
complexityis advice for the lane (large →luna:high).
3. Dedupe
gh issue list --state open --limit 200 --json number,title,body > open.json
helm jev check --preset dedupe --file dedupe-<slug>.json --jsonwhere dedupe-<slug>.json is {"candidate": "<draft text>", "against": [{"number", "title", "body"}, ...]} (Helm truncates each to 3,000 chars and runs one Jev call per pair).
Pass at most 40 open issues per call (the most recent, plus any whose title shares a word with the draft).
- P(same) ≥ 0.5: do not create. Comment on the existing issue with the deck link and any acceptance it lacks.
- "Related": create, and cross-link both issues.
4. Create
For each surviving draft:
gh issue create --title "<title>" --label <project label> --body-file drafts/<slug>.mdThen add - [ ] <feature> <url> to the map issue checklist (gh issue edit <map> --body-file ..., keeping the existing body), and fill Depends on with real issue numbers.
Rules
- Every created issue has passed
testable. No exceptions. - A duplicate feature produces no new issue.
- Without a Jev key the CLI returns
{ok:false, reason:'no key'}: stop and tell the supervisor; do not skip the checks silently.
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.

