Run any Skill in Manus with one click

standards-audit

Stars2

Forks1

UpdatedApril 11, 2026 at 21:25

Audits every loaded standard module for criterion counts, URL validity, level distribution, and equivalentTo reciprocal consistency. Catches silent drift between a standard's published version and our in-tree module.

Installation

Install with Codex or Claude Copy this prompt, paste it into Codex, Claude, or another assistant, and let it review the skill page and install it for you.

Run Skill in Manus

Source

vanctran

vanctran/ra11y

View GitHub Repository View Creator Repositories

Download

Run Skill in Manus

Related occupationsSOC

Based on SOC occupation classification

Software Quality Assurance Analysts and TestersComputer and Mathematical Occupations·SOC 15-1253

SKILL.md

readonly

name	standards-audit
description	Audits every loaded standard module for criterion counts, URL validity, level distribution, and equivalentTo reciprocal consistency. Catches silent drift between a standard's published version and our in-tree module.
allowed-tools	Bash(bun ) Bash(git )

/standards-audit

Deep audit of src/standards/ correctness.

Checks

Criterion counts match each standard's published total (wcag22: 86, wcag21: 78, section508 2017: the Chapter 5/6 counts, en301549 v3.2.1: the IT-specific counts).
Level distribution matches the spec (e.g., WCAG 2.2 has 30 A, 20 AA, 28 AAA — if these drift, our module is stale).
URL validity: every criterion.url resolves. Run scripts/check-docs-links.ts over the union.
equivalentTo reciprocal consistency: for every A.equivalentTo: [B], verify B exists in its standard. A reference to a removed criterion is a silent bug.
No duplicate local IDs within a standard.
No orphan criteria: every criterion is reachable from the standard's criteria array (guards against typos in the barrel).

Workflow

bun scripts/validate-standards.ts — runs all six checks and emits structured output.
If any fail: report specifically which standard, which criterion, and which check. Do not attempt to fix — this skill is read-only; a fix goes through standard-builder or type-smith.
If all pass: report "standards audit clean".

Return format

standards_audited: [wcag22, wcag21, section508, en301549]
checks:
  criterion_counts:      PASS
  level_distribution:    PASS
  url_validity:          PASS (302 URLs checked)
  reciprocal_equivalentTo: PASS
  duplicate_local_ids:   PASS
  orphan_criteria:       PASS
overall: PASS

More from this repository

same repository

continue

vanctran/ra11y

Parallel-track dispatcher. Reads .claude/backlog.md, picks the next unchecked item from each of up to 3 active tracks, dispatches them in parallel to specialist subagents, verifies, commits, and checks off. The primary entry point for autonomous ra11y development.

2026-05-052

add-standard

vanctran/ra11y

Scaffolds a new accessibility standard module (metadata, criteria, standard export, tests, kb entry). Delegates the criteria enumeration to spec-researcher + standard-builder. Use when adding any conformance framework.

2026-04-202

add-rule

vanctran/ra11y

Scaffolds a new ra11y rule end-to-end from a criterion ID using the evaluator-optimizer pattern — spec-researcher produces WCAG context, rule-implementer generates, a11y-reviewer critiques, iterate up to 3 times, then code-reviewer signs off. Use when adding any new rule.

2026-04-172

verify

vanctran/ra11y

Universal preflight — runs the full check sequence (typecheck, lint, tests, zero-dep, cycles, network isolation, limits, tsdoc, mermaid, kb drift) through the single `bun run verify` entrypoint. Use before every commit and before handing off to another agent.

2026-04-132

add-formatter

vanctran/ra11y

Scaffolds a new output formatter — file in src/output/formatters/, snapshot tests, CLI --format registration, and a short doc stub. Use when adding terminal/json/sarif/junit/html/markdown/plain or a custom format.

2026-04-112

bench

vanctran/ra11y

Runs scripts/bench.ts against the performance budget and reports the result. Fails if any scenario exceeds budget. Appends a row to docs/performance.md on success.

2026-04-112

name	standards-audit
description	Audits every loaded standard module for criterion counts, URL validity, level distribution, and equivalentTo reciprocal consistency. Catches silent drift between a standard's published version and our in-tree module.
allowed-tools	Bash(bun ) Bash(git )

/standards-audit

Deep audit of src/standards/ correctness.

Checks

Criterion counts match each standard's published total (wcag22: 86, wcag21: 78, section508 2017: the Chapter 5/6 counts, en301549 v3.2.1: the IT-specific counts).
Level distribution matches the spec (e.g., WCAG 2.2 has 30 A, 20 AA, 28 AAA — if these drift, our module is stale).
URL validity: every criterion.url resolves. Run scripts/check-docs-links.ts over the union.
equivalentTo reciprocal consistency: for every A.equivalentTo: [B], verify B exists in its standard. A reference to a removed criterion is a silent bug.
No duplicate local IDs within a standard.
No orphan criteria: every criterion is reachable from the standard's criteria array (guards against typos in the barrel).

Workflow

bun scripts/validate-standards.ts — runs all six checks and emits structured output.
If any fail: report specifically which standard, which criterion, and which check. Do not attempt to fix — this skill is read-only; a fix goes through standard-builder or type-smith.
If all pass: report "standards audit clean".

Return format

standards_audited: [wcag22, wcag21, section508, en301549]
checks:
  criterion_counts:      PASS
  level_distribution:    PASS
  url_validity:          PASS (302 URLs checked)
  reciprocal_equivalentTo: PASS
  duplicate_local_ids:   PASS
  orphan_criteria:       PASS
overall: PASS