Skip to main content
skill CX Codex

hermes-attestation-guardian — Claude Code skill

hermes-attestation-guardian — a skill that verifies Hermes infrastructure integrity at runtime. Generates signed attestations, compares them against baselines, and detects configuration drift.

AGPL-3.0 129 installs ★ 1,135 updated 2026-10-25

Install into Claude Code · Skills CLI (Vercel) All methods and platforms ↓
npx skills add prompt-security/clawsec --skill hermes-attestation-guardian -a claude-code
Open on GitHub

What hermes-attestation-guardian does

hermes-attestation-guardian produces deterministic posture snapshots of a Hermes server, cryptographically verifies them, and identifies configuration drift from an established baseline. It includes Node.js scripts for attestation generation, verification with fail-closed semantics, advisory feed updates, and recurring check scheduling.

Critical scope: works only with Hermes infrastructure (CLI, Gateway, profile-managed deployments). Requires node in PATH. For standalone setup, verify the release signature against ClawSec's public key before trusting the archive—the README contains a full verification procedure with SHA256 checks and OpenSSL validation.

The flow includes baseline comparison, advisory matching by severity level, and an emergency --allow-unsigned mode that must be disabled immediately after use.

What it helps with

  • verify Hermes server integrity and configuration at runtime
  • detect unexpected changes (drift) from baseline state
  • scan advisory feed and block vulnerable skill versions
  • schedule recurring attestation and advisory checks

Example prompts for the agent

  • Generate a snapshot of current Hermes posture: node scripts/generate_attestation.mjs
  • Verify drift against baseline with trusted checksum: node scripts/verify_attestation.mjs --baseline ~/.hermes/security/attestations/baseline.json --baseline-expected-sha256 <checksum>
  • Set up recurring advisory checks for a skill: node scripts/setup_advisory_check_cron.mjs --every 6h --skill some-skill --version 1.2.3 --apply

How to install hermes-attestation-guardian

1Skills CLI (Vercel)

npx skills add prompt-security/clawsec --skill hermes-attestation-guardian -a claude-code

2Manually: put the files in place

~/.claude/skills/hermes-attestation-guardian/SKILL.md

Copy the whole skill folder; for one project use .claude/skills/hermes-attestation-guardian/

Step by step, with commands for each platform: How to install skills in Claude Code: five ways

Questions about hermes-attestation-guardian

How do I install hermes-attestation-guardian in Claude Code?

Run npx skills add prompt-security/clawsec --skill hermes-attestation-guardian -a claude-code, then check the file at ~/.claude/skills/hermes-attestation-guardian/SKILL.md to confirm the installation.

What does hermes-attestation-guardian do?

It generates cryptographically signed snapshots of Hermes server state, verifies them against stored baselines to catch configuration drift, and manages advisory feed updates to block vulnerable skill versions.

What are the prerequisites?

node must be in PATH and you need a Hermes deployment (CLI or Gateway). For standalone installs, verify the release signature using the script in the README, which checks SHA256 hashes against ClawSec's public key.

Similar skills

All skills for Security →
azure-attestation MicrosoftDocs Reference skill for Azure Attestation work: error troubleshooting, claim-rule policies, SGX and TPM baselines, private endpoints. Keeps the agent on Microsoft's own docs. CX Codex +1 169installs MCP Server Attestation studiomeyer-io MCP Server Attestation validates tool manifests with Ed25519 signatures and polices spawn calls at runtime. Useful for shielding MCP setups against injected commands and rogue updates. CX Codex +1 Kenwea — Sandbox Attestation & Agent Marketplace kenwea-protocol Kenwea — Sandbox Attestation & Agent Marketplace is a Claude Code MCP server that checks npm packages and URLs in a sandbox and returns signed verdicts. No key or signup needed. CX Codex +1 ★ 2repo stars glab-attestation vince-winkintel The glab-attestation Claude Code skill works with GitLab attestations and checks software artifact provenance. It activates for requests about artifact verification or supply chain security. CX Codex +1 113installs hermes-history-ingest Ar9av hermes-history-ingest — a skill for Claude Code, Codex, Cursor and other platforms that turns Hermes agent history into Obsidian wiki pages. Extracts patterns from memory files and session logs, groups them by topic and project. CX Codex +1 2,500installs hermes-voice-setup hermes-voice contributors The hermes-voice-setup Claude Code skill connects a Hermes profile to Home Assistant Assist for voice conversations. It sets up the profile and integration for an Echo or another Assist voice device. CX Codex +1 ★ 732repo stars

From the same set9

clawsec-suite The clawsec-suite skill monitors ClawSec advisories, verifies signatures, and checks installed skills for affected entries. Use it when security checks need approval gates. CX Codex +1 881installs openclaw-audit-watchdog The openclaw-audit-watchdog skill runs a deep security audit of OpenClaw agents on a schedule and sends the report as a direct message. Email delivery is optional. CX Codex +1 756installs clawsec-clawhub-checker clawsec-clawhub-checker — a skill that vets code before installation by checking ClawHub reputation scores, scanner status, and authorship history. Install it if you want an extra approval layer for skill installs across your environments. CX Codex +1 259installs soul-guardian The soul-guardian skill watches agent workspace files against an approved baseline, restores selected files, and raises alerts when drift appears. CX Codex +1 258installs clawsec-feed clawsec-feed provides a consolidated feed of OpenClaw security advisories and threats. Use it when an agent needs current vulnerability context. CX Codex +1 243installs clawtributor clawtributor enables agents to document prompt attacks, sandbox evasions, and dangerous extensions for collective security feeds. CX Codex +1 208installs clawsec-scanner The clawsec-scanner skill checks project dependencies and source code for security issues, and inspects OpenClaw hooks. Use it for a project review or recurring scans. CX Codex +1 195installs clawsec-nanoclaw clawsec-nanoclaw — a skill that monitors vulnerabilities in installed skills before adding them or during regular audits. Guards against compromised dependencies using a signed advisory feed. CX Codex +1 189installs

Project sponsors

Companies that keep this analytics open