agentleFS
omarismailbomarismailb/product-wiki/.agents/skills/propose-change/SKILL.md

propose-change

Skill8 starsChanged 5 months agoMITView on GitHub

Use when a request may change product behaviour, fix a bug, alter a workflow, add a capability, clarify product intent, or create/refine requirements before code.

1 other project uses a version of this.

Adopted by

Most recently taken up by

  • Other versions1 project uses a version of this.

    Word for word · 0

    None found.

    Edited · 0

    None found.

    Wrote their own version · 1

    1. Vladick-Pick/business-ontology★ 3same name

    How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.

  • In this repository8 other files it gives its agents.

The file

Show its rules and text

We split this skill into the rules it gives an agent

16 rules. About 0 are common, 12 are distinctive, and 4 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈12

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 4

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Do not edit application code.

    line 10
  2. Distinctive:

    Do not invoke implementation, scaffold application code, or edit application code until the proposal is approved.

    line 14
  3. Distinctive:

    Before drafting, load these files:

    line 20
  4. Very specific:

    If either file is missing, do not infer the proposal shape from schemas or previous examples.

    line 25
  5. Distinctive:

    Stop only if the canonical contract cannot be restored.

    line 28
  6. Very specific:

    Read AGENTS.md, wiki/overview.md, wiki/index.md, relevant wiki units, and recent proposals.

    line 32
  7. Very specific:

    Read templates/proposal-template.md and references/proposal-rubric.md.

    line 33
  8. Distinctive:

    Ask one question at a time when an answer would materially change scope, user experience, data, security, architecture, or checks.

    line 54
  9. Distinctive:

    Do not ask preference questions that do not affect the proposal.

    line 56
  10. Distinctive:

    Keep every required section, even when the answer is "none", "not affected", or "deferred".

    line 61
  11. Distinctive:

    no hidden implementation decision

    line 64
  12. Distinctive:

    no duplicate capability without explanation

    line 65
  13. Distinctive:

    all acceptance criteria are observable

    line 66
  14. Very specific:

    Run node scripts/proposal-lint.mjs and node scripts/proposal-traceability-lint.mjs.

    line 72
  15. Distinctive:

    Stop and ask the user to approve, edit, or reject it.

    line 73
  16. Distinctive:

    Create a full plan when the request is a tiny mechanical edit.

    line 98

≈ “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.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Propose Change

Your job is to turn a normal product request into a clear product wiki proposal. Stay in natural language. Do not edit application code.

Hard gate

Do not invoke implementation, scaffold application code, or edit application code until the proposal is approved. This applies even when the request sounds simple. The proposal can be short, but the gate still exists.

Template contract

Before drafting, load these files:

  • templates/proposal-template.md
  • .agents/skills/propose-change/references/proposal-rubric.md

If either file is missing, do not infer the proposal shape from schemas or previous examples. First run node scripts/repair-contracts.mjs --write, then re-run node scripts/template-lint.mjs and node scripts/skill-lint.mjs. Continue if repair succeeds. Stop only if the canonical contract cannot be restored.

Workflow

  1. Read AGENTS.md, wiki/overview.md, wiki/index.md, relevant wiki units, and recent proposals.
  2. Read templates/proposal-template.md and references/proposal-rubric.md.
  3. Classify the request:
    • mechanical: copy, styling, rename, or config-only with no product intent change.
    • normal: feature, bug, workflow, or user-facing behaviour change.
    • high-risk: auth, money, privacy, data, migration, external integrations, architecture, design system, or multiple modules.
    • greenfield: first product slice in an empty wiki or new app.
    • brownfield: existing codebase import or retrofit. Use import-codebase instead.
  4. Restate the request in one plain sentence.
  5. Build a product-unit coverage map before asking questions:
    • actor
    • job
    • story
    • journey
    • capability
    • rule
    • acceptance criterion
    • outcome
    • non-goal
    • assumption or risk
    • glossary term
    • decision
  6. Ask one question at a time when an answer would materially change scope, user experience, data, security, architecture, or checks. Recommend a default answer when the repo context gives you one. Do not ask preference questions that do not affect the proposal.
  7. For meaningful changes, present 2-3 product or architecture approaches with trade-offs and your recommendation. Skip this only for truly mechanical changes.
  8. Identify the smallest useful wiki change and explain how it fits the whole-product picture in wiki/overview.md.
  9. Draft the proposal in intake/proposals/ from templates/proposal-template.md. Keep every required section, even when the answer is "none", "not affected", or "deferred".
  10. Mark the proposal status: awaiting-approval.
  11. Self-review the proposal against the rubric:
    • no hidden implementation decision
    • no duplicate capability without explanation
    • all acceptance criteria are observable
    • checks can run against code later
    • explicit out-of-scope items
    • confidence and assumptions are labelled
    • each distinct user-requested action is its own story, not collapsed into one (see the rubric's "Sizing and splitting units")
    • real trade-offs are decisions with rejected alternatives, not buried in rules
  12. Run node scripts/proposal-lint.mjs and node scripts/proposal-traceability-lint.mjs.
  13. Stop and ask the user to approve, edit, or reject it.

Proposal must include

  • Request.
  • Why now.
  • Request type and risk level.
  • Affected actors, jobs, stories, journeys, capabilities, rules, acceptance criteria, decisions, outcomes, non-goals, assumptions, risks, and glossary terms.
  • Proposed wiki additions or updates.
  • Acceptance criteria with stable IDs.
  • Checks that should run against the code later.
  • Reuse or refactor question.
  • Design-system or architecture impact when relevant.
  • Open decisions that need human judgement.
  • Explicit out-of-scope items.
  • Confidence level for inferred units.
  • Self-review notes.

Do not

  • Write implementation code.
  • Rewrite large parts of the wiki without approval.
  • Hide product decisions inside implementation detail.
  • Collapse jobs, stories, and acceptance criteria into one vague requirement.
  • Skip a unit because it feels obvious.
  • Create a full plan when the request is a tiny mechanical edit.

Related skills

  • apply-wiki-change: after the proposal is approved.
  • compile-change: after the approved wiki change exists.
  • import-codebase: when the request is to retrofit an existing repo instead of proposing one feature.

Download now (ZIP)

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
What did you use it for? e.g. Reviewing pull requests in a Rails app

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.