spec
/spec
Write a spec to the owning workspace's specs/ folder from what the session already settled, with no new interview. Problem first, testable success criteria, and the seams to test. Use for /spec, or when someone says spec this out, write the plan, or turn this thread into a spec.
- Author
- foxrun-ai ↗
- Repository
- foxrun ↗
- Licence
- MIT
- Last repository change
- 26.10.2026
Assessment and security
- Files in this package
- 3
- Scripts
- 0
- HTML comments
- 0
- URLs in SKILL.md
- 0
No signals detected. Repository changed on 26.10.2026, without references to retired models or APIs.
MIT, open: use and modify while retaining author attribution.
What it does and when to use it
Write a spec to the owning workspace's specs/ folder from what the session already settled, with no new interview. Problem first, testable success criteria, and the seams to test. Use for /spec, or when someone says spec this out, write the plan, or turn this thread into a spec.
Arguments
/spec <name> [workspace] [task id]
Original author description
Write a spec to the owning workspace's specs/ folder from what the session already settled, with no new interview. Problem first, testable success criteria, and the seams to test. Use for /spec, or when someone says spec this out, write the plan, or turn this thread into a spec.
How it works
- Steps
- Gotchas
SKILL.md preview
--- type: note workspace: system timestamp: 2026-10-26T14:25:00-04:00 name: spec description: "Write a spec to the owning workspace's specs/ folder from what the session already settled, with no new interview. Problem first, testable success criteria, and the seams to test. Use for /spec, or when someone says spec this out, write the plan, or turn this thread into a spec." kind: knowledge argument-hint: "<name> [workspace] [task id]" --- # spec A spec turns a settled thread into a plan another session can build cold. Write from what the session already knows. Do not interview the person again; if a question is still open, grill first. A spec lives in `~/foxrun/<ws>/specs/` and names its task, so the work and the plan point at each other. ## Steps 1. **Find the home.** Set `<ws>` to the workspace that owns the work. Check that `~/foxrun/<ws>/AGENTS.md` exists, and read it. If the workspace does not exist, stop: do not guess a folder. Done when you read the workspace's `AGENTS.md`. 2. **Read the ground.** Read the code, pages, rulings, and rows the spec touches. Run `python3 ~/.foxrun/foxrun.py rulings <words>`. Use the glossary words. Done when each ruling that touches the work is named. 3. **Find the seams.** Name the places where a test can prove the work, highest first. Prefer one seam that exists. Show the seams to the person in a few lines. Done when the person agreed the seams, or you noted they were not asked. 4. **Draft.** Follow [references/template.md](references/template.md). Match depth to risk: a cleanup gets a short spec; a data move or a delete gets every section. Name the file `<task id>-<kebab-slug>.md` when a task exists. Done when every required section has content and each success criterion says how it is measured. 5. **Check, then write.** Run `python3 ~/.foxrun/base/lint_gate.py --mode write --target ~/foxrun/<ws>/specs/<name>.md --body <draft file>`. Done when the gate passed and the spec is at its path. 6. **Tie it to the task.** `python3 ~/.foxrun/work/work.py --workspace <ws> --db <tracker> --receipt-root ~/foxrun/<ws> --by <you> comment <task id> "spec: specs/<name>.md"`. Done when the comment prints `ok`, or you said no task exists yet. ## Gotchas - No problem means no spec. Lead with what breaks and the evidence. - Leave out file paths and code that will go stale, unless a prototype settled a shape that prose cannot hold. - A spec is a plan, not a go. Building it needs its own task and, for a one-way door, the person's yes.
Files in this package (3)
- SKILL.md 2 KB
- references/template.md 1 KB
- replays/no-interview.json 1 KB
Versions
| Repository | Score | Licence |
|---|---|---|
| This page · foxrun-ai/foxrun | 90 | MIT |
| renzrollon/interlock | 100 | MIT |
| CodaColaLight/nightmill | 100 | Apache-2.0 |
| duronext/jig | 100 | MIT |
| optiflowic/keyturn | 100 | MIT |
| rjmurillo/ai-agents | 100 | MIT |
| rjmurillo/ai-agents | 100 | MIT |
| rjmurillo/ai-agents | 100 | MIT |
| flavien-ia/hypervibe-craft | 100 | Apache-2.0 |
Database snapshot: 13.11.2026 · package version 21ff2e5de326 (SHA-256 digest of the entire package)
Downloads in the last 30 days: 0 · ranking
Report a problem with this skill: harmful content, incorrect descriptions or an author’s removal request.
Similar skills
- 90writing-product-specs
Use when asked to write a product spec, PRD, or requirements document, or to design a new feature or project before implementation - interactively elicits problem, audience, success criteria, and scope, then produces a structured spec file - 95to-spec
Turn the current conversation into a spec and publish it to the project issue tracker: no interview, just synthesis of what you've already discussed. Use when the conversation is ready to become a written spec. Triggers on: to-spec, to-prd, write this up as a spec. - 100to-feature-spec
Turns the current conversation and codebase understanding into a feature spec. Use when asked to capture agreed requirements, write a feature spec, or turn the current context into a PRD; not to start a new requirements interview or implementation. - 100to-spec
Turns the current conversation into a spec and publishes it to the project issue tracker, with no interview. Use when the user says \"write a spec\", \"create a PRD\", or wants the discussion captured as a spec. - 100write-spec
Turns a vague feature idea or problem statement into a structured, review-ready feature spec or PRD with goals, non-goals, success metrics, and testable acceptance criteria. Use when a PM says "write a PRD", "spec this feature", "turn this idea into a spec", "scope this",… - 100to-spec
Turn the current conversation into a spec and publish it to the project issue tracker: no interview, just synthesis of what you've already discussed. Use when the user types /spec or explicitly asks for this step; do not activate on your own for unrelated requests.