Skip to content

docs(examples): align ai-register manifest with EU AI Act, NIST RMF, ISO 42001#1184

Merged
christso merged 1 commit intomainfrom
docs/ai-register-governance-fields
Apr 28, 2026
Merged

docs(examples): align ai-register manifest with EU AI Act, NIST RMF, ISO 42001#1184
christso merged 1 commit intomainfrom
docs/ai-register-governance-fields

Conversation

@christso
Copy link
Copy Markdown
Collaborator

Summary

Closes gaps identified in a governance review of the .ai-register.yaml example against EU AI Act Annex IV, NIST AI RMF, ISO 42001, and CycloneDX-ML conventions.

Added fields:

  • version — required by EU AI Act Annex IV Art. 11, ISO 42001 A.9, CycloneDX-ML
  • contact — compliance/incident contact; EU AI Act Art. 8, NIST RMF GOVERN-2
  • created_at — lifecycle tracking alongside last_reviewed; ISO 42001 A.9
  • intended_use — authorised use contexts, distinct from general description; EU AI Act Annex IV §1(a), NIST RMF GOVERN-1.1
  • limitations — known failure modes and out-of-scope uses; NIST RMF MEASURE-1.1, EU AI Act Annex IV §2
  • human_oversight — override authority and escalation triggers for risk_tier: high; EU AI Act Art. 14, NIST RMF GOVERN-4
  • EU-AI-ACT-2024:Art.14 added to controls list (now documented)

Removed YAGNI fields:

  • evals.runs_in_ci — no documented consumer; creates a second source of truth vs. CI YAML config
  • deployment — no governance framework uses it; implied by the service repo's deployment pipeline

README: Added a field reference table mapping each field to the standard that requires it.

Test plan

  • Pre-push hook: Build, Typecheck, Lint, Test, Validate eval YAML — all passed
  • Visual review of the example YAML reads naturally as a governance artifact

…ISO 42001

Add missing fields required by the governance frameworks the manifest cites:
- version, contact, created_at (lifecycle tracking; EU AI Act Annex IV Art. 11)
- intended_use (distinct from description; EU AI Act Annex IV §1(a), NIST RMF GOVERN-1.1)
- limitations (NIST RMF MEASURE-1.1, EU AI Act Annex IV §2)
- human_oversight (required for risk_tier: high; EU AI Act Art. 14, NIST RMF GOVERN-4)
- EU-AI-ACT-2024:Art.14 to controls list

Remove YAGNI fields with no documented consumer:
- evals.runs_in_ci (creates a second source of truth vs CI config; nothing reads it)
- deployment (no governance framework uses it; implied by the service repo context)

Add field reference table to README documenting which standard requires each field.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
@cloudflare-workers-and-pages
Copy link
Copy Markdown

Deploying agentv with  Cloudflare Pages  Cloudflare Pages

Latest commit: 057025e
Status: ✅  Deploy successful!
Preview URL: https://04f3cfaf.agentv.pages.dev
Branch Preview URL: https://docs-ai-register-governance.agentv.pages.dev

View logs

@christso christso marked this pull request as ready for review April 28, 2026 01:00
@christso christso merged commit d670b47 into main Apr 28, 2026
4 checks passed
@christso christso deleted the docs/ai-register-governance-fields branch April 28, 2026 01:01
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant