项目文件夹

文件

T

tommylauren 7d7fb736be feat: add signed-audit-trails teaching plugin

Companion to the existing protect-mcp and review-agent-governance plugins
in the governance category. This one is a teaching skill, not a runtime
hook: a cookbook-style walkthrough for explaining, evaluating, or
demonstrating the signed-audit-trails pattern before committing to the
protect-mcp infrastructure.

Contents
────────
- plugin.json, README.md (explains this is a teaching skill, not runtime)
- skills/signed-audit-trails-recipe/SKILL.md (~1600 words) walking through:
1. Hook configuration in .claude/settings.json
2. Cedar policy authoring (four permit/forbid rules)
3. Using Claude Code normally with hooks active
4. Inspecting a produced receipt
5. Verifying the chain with @veritasacta/verify
6. Demonstrating tamper detection live
Plus: three-invariant cryptographic model (JCS + Ed25519 + hash chain),
cross-implementation interop table (4 implementations), CI/CD YAML
snippet, SLSA composition via ResourceDescriptor byproduct, common
pitfalls, references.

Marketplace entry
─────────────────
Added under category: "governance" immediately after protect-mcp. Keywords
emphasize the tutorial/cookbook nature ("tutorial", "skill", "recipe") to
distinguish from the runtime plugins.

Why this is useful alongside the two existing plugins
─────────────────────────────────────────────────────
protect-mcp is runtime. review-agent-governance is runtime (with human-
approval pattern). signed-audit-trails is education. A user evaluating
whether to adopt receipts can invoke the skill to get the concept and a
reproducible demo in-session, then install the runtime plugin if they
decide to proceed. Pairs naturally with /audit-chain and /verify-receipt
slash commands shipped in protect-mcp.

No new dependencies. Same IETF draft + @veritasacta/verify references as
the other governance plugins.

2026-04-17 12:48:11 -04:00

2.1 KiB

原始文件 永久链接 Blame 文件历史

signed-audit-trails

A teaching skill for setting up cryptographically signed audit trails on every Claude Code tool call. Cookbook-style walkthrough with runnable examples.

What this is

A skill (not a runtime hook): a set of instructions and examples that explain the pattern end-to-end. Use this when you are figuring out whether receipts are the right fit for your project. Once you know they are, install the protect-mcp plugin for the actual hooks.

When to use this plugin

Learning the pattern before committing to infrastructure
Evaluating whether signed audit trails fit your compliance need
Teaching team members the three-invariant cryptographic model (JCS canonicalization + Ed25519 signatures + hash chains)
Walking a client or auditor through a live demonstration of tamper detection

For production use, the protect-mcp plugin gives you the runtime hooks directly. This plugin is the skill file you invoke via Skill when you want the concept explained in-session.

What is inside

skills/signed-audit-trails-recipe/SKILL.md

A single skill file containing:

Step-by-step setup (Cedar policy, hook configuration, first receipt)
Live tamper detection walkthrough
Receipt format explanation (three invariants)
Cross-implementation interoperability table
CI/CD integration snippet (GitHub Actions)
Composition with SLSA provenance for agent-built software
Common pitfalls and references

Standards

Ed25519 (RFC 8032) for receipt signatures
JCS (RFC 8785) for deterministic JSON canonicalization before signing
Cedar (AWS) for policy evaluation
IETF draft draft-farley-acta-signed-receipts

protect-mcp — the runtime hook implementation
review-agent-governance — require human approval before review-surface actions; composes with protect-mcp

License

MIT. Same as the adjacent governance-category plugins in this marketplace.

2.1 KiB 原始文件 永久链接 Blame 文件历史