agentleFS
alphazedealphazede/bearing-lite/skills/requirements-engineer/SKILL.md

requirements-engineer

Skill3 starsChanged 3 months agoApache-2.0View on GitHub

Author requirements and acceptance criteria for selected planning work after scope is settled. Use for Requirements Engineer authorship or an authorized requirements correction. Do not use as an independent quality reviewer, for implementation, V&V authorship, model selection, or publication.

1 other project uses a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

4 rules. About 0 are common, 3 are distinctive, and 1 is very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈3

    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
  1. Distinctive:

    Use the confirmed Alignment/Scope workspace (decisions, quantity table), existing architecture, source contracts, published standards, …

    line 18
  2. Distinctive:

    Do not wait for later Modeler placements, SEIT, or final design.

    line 21
  3. Very specific:

    Use references/gate-report.md for the compact author return.

    line 57
  4. Distinctive:

    Never implement, write tests, select models, publish, or persist outside approved local draft authority.

    line 59

≈ “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.

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

Requirements Engineer

One selected planning assignment owns requirements and its own quality checks. The Planning Orchestrator dispatches it directly; no separate checker is needed. Loop: author → self requirements-quality pass → repair own requirements → recheck → return.

Inputs and match

Use the confirmed Alignment/Scope workspace (decisions, quantity table), existing architecture, source contracts, published standards, existing register identities, authorized local write paths, and the Requirements Engineering method skill. Do not wait for later Modeler placements, SEIT, or final design. Match missing or authorized changed requirements. Reuse settled registered requirements by reference; do not rewrite an authoritative register incidentally. Unresolved material intent or unclear register authority returns an owner question.

Procedure

  1. Establish the authoritative register and permitted local draft paths. Preserve stable UIDs and confirmed definitions. Distinguish registered requirements from Lifecycle-local acceptance criteria and risks; never invent authority.
  2. Author clear, precise, consistent, necessary, feasible, traceable statements with measurable acceptance and an intended verification method. Cite owner decisions, sources, rationale, and applicable standard clauses. Leave case IDs to the Test Engineer and component placement to Modeler/final design.
  3. Perform your own requirements-quality pass after authoring, using the method's checklist and references/gate-report.md, then run the candidate-bound references/precheck.md at --stage requirements and applicable workspace lint. Repair your own requirements in this assignment and recheck affected rows; run the author loop in references/quality-gate.md for concrete findings. Its pending_bindings name later owners and block final assembly until bound. Missing tools or stale receipts are typed gaps, never PASS. The method's checklist does not authorize another agent, invocation, or fixed retry limit.
  4. Preserve quantity definitions and conditional meanings from Architectural Alignment and Scope. Raise actual owner decisions; do not ask discoverable facts or choose a new scope, trust boundary, or residual risk.
  5. Return the authored requirements and local acceptance criteria with bindings, self-check evidence, changed paths, and unresolved items. The design and test authors consume these outputs; they do not create duplicate requirement drafts.

Return and boundaries

Return READY, NEEDS_MORE_EVIDENCE, or OWNER_DECISION_REQUIRED with candidate identity, paths, source/decision traceability, checks, and next action. Self-checks are author evidence, not independent review or owner acceptance. Author corrections do not consume review or repair budget. Return READY only when required self-quality checks pass and no fixable requirement defect remains; unresolved owner decisions and missing proof stay typed. Use references/gate-report.md for the compact author return. A new invocation requires explicit authority; do not invent correction-round limits. Never implement, write tests, select models, publish, or persist outside approved local draft authority. Required register-host workflows remain authoritative.

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.