Skip to main content
AI/MLsickn33

akf-trust-metadata

The AI native file format. EXIF for AI — stamps every file with trust scores, source provenance, and compliance metadata. Embeds into 20+ formats (DOCX, PDF, images, code). EU AI Act, SOX, HIPAA auditing.

Stars
39,227
Source
sickn33/antigravity-awesome-skills
Updated
2026-05-30
Slug
sickn33--antigravity-awesome-skills--akf-trust-metadata
View on GitHubRaw SKILL.md

// install — copy + paste into any project

mkdir -p .claude/skills && curl -fsSL https://raw.githubusercontent.com/sickn33/antigravity-awesome-skills/HEAD/plugins/antigravity-awesome-skills-claude/skills/akf-trust-metadata/SKILL.md -o .claude/skills/akf-trust-metadata.md

Drops the SKILL.md into .claude/skills/akf-trust-metadata.md. Works with Claude Code, Cursor, and any agent that loads SKILL.md files from .claude/skills/.

AKF — The AI Native File Format

Every photo has EXIF. Every song has ID3. AKF is the native metadata format for AI-generated content.

When to Use

Use this skill when you need to stamp, inspect, or audit provenance and trust metadata on AI-generated or AI-modified files for compliance, review, or handoff workflows.

After creating or modifying files

akf stamp <file> --agent <agent-name> --evidence "<what you did>"

Evidence examples:

  • "generated from user prompt"
  • "refactored existing code"
  • "tests pass"
  • "docs reviewed"

Before modifying existing files

akf read <file>          # Check existing trust metadata
akf inspect <file>       # See detailed trust scores

Compliance auditing

akf audit <file> --regulation eu_ai_act    # EU AI Act Article 50
akf audit <file> --regulation hipaa         # HIPAA
akf audit <file> --regulation sox           # SOX
akf audit <file> --regulation nist_ai       # NIST AI RMF

Classification

  • --label confidential for finance/secret/internal paths
  • --label public for README, docs, examples
  • Default: internal

Install

pip install akf

Links

Limitations

  • Use this skill only when the task clearly matches the scope described above.
  • Do not treat the output as a substitute for environment-specific validation, testing, or expert review.
  • Stop and ask for clarification if required inputs, permissions, safety boundaries, or success criteria are missing.