severity-mapping
Skill409 starsChanged 6 months agoMITView on GitHub
Canonical severity level definitions and cross-domain mapping for web, document, and markdown audits. Score impact ranges, WCAG conformance alignment, and cross-format normalization.
In this repository31 other files it gives its agents.
- AGENTS.mdAGENTS.md
- .claude/AGENTS.mdAGENTS.md
- claude-code-plugin/AGENTS.mdAGENTS.md
- claude-code-plugin/CLAUDE.mdCLAUDE.md
- CLAUDE.mdCLAUDE.md
- .github/copilot-instructions.mdCopilot instructions
- accessibility-rulesSkill
- ci-integrationSkill
- cognitive-accessibilitySkill
- data-visualization-accessibilitySkill
- design-systemSkill
- document-scanningSkill
- email-accessibilitySkill
- framework-accessibilitySkill
- github-a11y-scannerSkill
- github-analytics-scoringSkill
- github-scanningSkill
- github-workflow-standardsSkill
- help-url-referenceSkill
- legal-compliance-mappingSkill
- lighthouse-scannerSkill
- markdown-accessibilitySkill
- media-accessibilitySkill
- mobile-accessibilitySkill
- office-remediationSkill
- playwright-testingSkill
- python-developmentSkill
- report-generationSkill
- testing-strategySkill
- web-scanningSkill
- web-severity-scoringSkill
The file
We split this skill into the rules it gives an agent
3 rules. About 0 are common, 3 are distinctive, and 0 are 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.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 38
When generating a unified report that combines web, document, and markdown findings:
- Distinctive: line 43
Do not double-count a finding that appears in both an axe-core scan and a Playwright behavioral scan -- deduplicate by rule ID and element …
- Distinctive: line 57
Document and markdown audits do not use confidence multipliers -- they use the base score impact only.
≈ “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.
Files
Severity Mapping
Canonical severity definitions shared across web, document, and markdown audit domains.
Canonical Severity Levels
| Level | Definition | WCAG Conformance | Score Impact |
| Critical | Blocks access entirely for one or more user groups. No workaround exists. | Fails Level A | -15 to -22 points |
| Serious | Severely impairs access; a workaround may exist but is unreasonable. | Fails Level A or AA | -5 to -10 points |
| Moderate | Causes friction or confusion; most users can still complete the task. | Fails Level AA (advisory) | -2 to -4 points |
| Minor | Polish issue; no meaningful barrier for most users. | Technique deviation | -1 point |
Cross-Domain Equivalents
| Canonical Level | Web (axe-core / Playwright) | Document (Office/PDF) | Markdown |
| Critical | critical (axe-core) | Missing document title; untagged PDF; missing heading structure | Broken anchor link leading to dead navigation |
| Serious | serious (axe-core) | Missing alt text on informational image; table has no headers | Ambiguous link text (e.g., "click here"); missing alt text on image |
| Moderate | moderate (axe-core) | Merged table cells; color-only data encoding | Skipped heading level; missing table caption |
| Minor | minor (axe-core) | Decorative image with non-empty alt; redundant bookmark | Emoji in heading; em-dash used as list separator |
Score Impact Rules
Score impacts listed above are additive per finding. When the same issue recurs across many elements:
- First occurrence — full deduction
- Same issue type, same page/file — no additional deduction (already counted once per issue rule)
- Same issue type, different page/file — full deduction applies again to the new page/file score
Conversion Rules (Normalizing Cross-Domain Reports)
When generating a unified report that combines web, document, and markdown findings:
- Map each finding to the canonical severity level using the table above.
- Apply the score impact from the canonical level, not from the domain-specific tool's internal scoring.
- For aggregated scores (portfolio or project level), compute each page/file score independently, then average.
- Do not double-count a finding that appears in both an axe-core scan and a Playwright behavioral scan -- deduplicate by rule ID and element location.
Confidence Modifier (Web Domain)
For web audits, confidence level modifies the base deduction:
| Confidence | Multiplier | Source Signals |
| Confirmed | 1.2x | axe-core + agent code review + Playwright behavioral test |
| High (two sources) | 1.0x | Any two of the three sources above |
| High (one source) | 1.0x (reduced base) | Single automated tool, high rule reliability |
| Medium | 0.7x | Single source, medium rule reliability |
| Low | 0.3x | Single source, low rule reliability or heuristic |
Document and markdown audits do not use confidence multipliers -- they use the base score impact only.
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
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.