agentleFS
nesnilnehcnesnilnehc/ai-cortex/skills/review-requirements/SKILL.md

review-requirements

Skill8 starsChanged 2 months agoMITView on GitHub

Review an existing requirement document against its modeling Spec and canonical quality Rule before design begins. Evaluative atomic skill; outputs findings without rewriting.

3 other projects use a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

8 rules. 0 are common, 7 are distinctive, and 1 is very specific to this project.

  • 0

    Common

    Most of the 3 edited versions we read keep these.

  • 7

    Distinctive

    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:

    After a requirement changes and downstream impact must be re-evaluated

    line 27
  2. Distinctive:

    Before adopting an externally authored requirement into the repository

    line 28
  3. Distinctive:

    Verify the artifact type, frontmatter contract and required/conditional sections against the Spec.

    line 33
  4. Distinctive:

    If no findings remain, state that the artifact is ready for the next design layer.

    line 37
  5. Distinctive:

    Do not invent requirements, metrics, risks or acceptance criteria.

    line 51
  6. Very specific:

    Do not use the obsolete generic R-NN model when the active Spec defines project requirement IDs and document sections.

    line 52
  7. Distinctive:

    Do not mix solution architecture into requirement findings.

    line 53
  8. Distinctive:

    If Scope is inapplicable and omitted, do not emit an ordering finding for its absence.

    line 93

“Common” is measured: kept by at least 6 in 10 of the 3 edited versions of this file we could read. “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

Skill: Review Requirements

Purpose

Evaluate an existing requirement against requirement-modeling and requirement-quality. Emit findings-list output; do not author or rewrite the requirement.

Core objective

Decide whether the requirement is a safe, testable and traceable input to design, including measurable quality-attribute scenarios when their trigger applies.

Scope boundaries

This Skill reviews one requirement artifact. It does not elicit needs, design a solution, review code, approve open decisions or repair the document.

Use cases

  • Immediately after a requirement draft, before functional or technical design
  • After a requirement changes and downstream impact must be re-evaluated
  • Before adopting an externally authored requirement into the repository

Behavior

  1. Load the modeling Spec and quality Rule in full.
  2. Verify the artifact type, frontmatter contract and required/conditional sections against the Spec. Check retained body sections against the relative order in Spec §5.2.1, allowing omission only according to section applicability. Check status-dependent metadata as structural compliance, not as a content-quality exemption.
  3. Apply every quality criterion to the document and its cited local parent where traceability requires it.
  4. Apply the Rule's unresolved-decision criteria across the whole document, including placeholders, assumptions and deferred acceptance targets. Verify the actual effect of each open item rather than trusting its blocking/non-blocking label. Distinguish missing requirement decisions from design choices that remain within established constraints.
  5. Emit one finding per violation with category requirement-quality, precise section/criterion location and, when useful, an actionable suggestion.
  6. If no findings remain, state that the artifact is ready for the next design layer; do not create that design.

Input and output

Input is one existing requirement path or its full content. Output is a findings list. Zero findings is a gate result, not a score.

The formal requirement template is an authoring aid. Review against the active Spec and Rule, not a literal template match; placeholders and matching headings do not establish compliance. The capture-work-items requirement template produces a backlog item, which is not the formal requirement artifact reviewed here.

Restrictions

  • Do not invent requirements, metrics, risks or acceptance criteria.
  • Do not use the obsolete generic R-NN model when the active Spec defines project requirement IDs and document sections.
  • Do not mix solution architecture into requirement findings.

Self-Check

  • The active Spec and Rule were loaded.
  • Required and conditionally required sections were evaluated, and retained sections follow the Spec’s relative order.
  • Quality scenarios were checked when the trigger applied.
  • The same content criteria were applied regardless of lifecycle status.
  • Unresolved decisions were checked throughout the document; non-blocking labels and deferred design choices were verified against their actual impact.
  • Every finding is document-grounded, precise and category-compliant.
  • The requirement was not rewritten.

Examples

Example 1: non-functional target is vague

Input: “The API should be fast.” Expected: a major finding at the acceptance section requiring a measurable target and representative conditions; do not choose the number.

Example 2: small functional change

Input: a bounded UI wording change with no architecture-significant quality attribute. Expected: do not manufacture a Quality Attribute Scenarios section; review the normal requirement contract.

Example 3: unresolved business rule under different statuses

Input: eligibility is "to be decided", marked non-blocking, in either a draft or an approved requirement. Expected: the same major finding because eligibility and its acceptance result are indeterminate. An owner and a resolution date do not resolve the rule; do not choose the rule on the user's behalf.

Example 4: deferred implementation choice

Input: the requirement establishes retention duration, deletion behavior and verifiable acceptance conditions, while the storage engine is left to technical design. Expected: do not report the implementation choice itself as an unresolved requirement decision. Check that the established constraints remain sufficient.

Example 5: section order

Input: Scope appears after Acceptance Criteria. Expected: a minor finding identifying the misplaced section and its required position under Spec §5.2.1. If Scope is inapplicable and omitted, do not emit an ordering finding for its absence. Evaluate missing applicable sections and content defects separately.

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.