Skip to content
Back to search

clawsec

MarketplaceWorks with Claude CodeGitHub

A complete security skill suite for OpenClaw, Hermes, PicoClaw and NanoClaw agents (and variants). Protect your SOUL.md (etc') with drift detection, live security recommendations, automated audits, and skill integrity verification. All from one installable suite.

Safety scan

A static text scan, not a guarantee. Select the badge to see what was scanned and every finding.

Checked today

Check it deeper on your machine

It has changed since its scan on 11 Nov 2026.

As a Claude Code plugin, install it and then ask Claude Code:

/plugin marketplace add keithmackay/mackayi/plugin install scanskillsafety@mackayiDo a deep scan of https://github.com/prompt-security/clawsec with scanskillsafety

Or from a terminal, clone it and run it:

git clone https://github.com/keithmackay/scanskillsafety-skill ~/.claude/skills/scanskillsafetynode ~/.claude/skills/scanskillsafety/cli.cjs --deep 'https://github.com/prompt-security/clawsec'

It checks the exact commit: every file, its packages against OSV (api.osv.dev), how it compares with this page's scan (findsafeskills.com receives the repository's address), and, if you agree, Claude reading every file with your own Claude subscription. It never runs the code.

Badge for your README
findsafeskills safety rating

Markdown

[![findsafeskills safety rating](/api/listings/prompt-security--clawsec/badge.svg)](/listings/prompt-security--clawsec)

HTML

<a href="/listings/prompt-security--clawsec"><img src="/api/listings/prompt-security--clawsec/badge.svg" alt="findsafeskills safety rating" /></a>

Install

Install instructions from the README

As written by the repo's author. findsafeskills has not verified these steps.

The OpenClaw entry point is `clawsec-suite`. Adding the package and activating its persistent hook are separate, reviewable steps.

1. Add the suite

npx skills add prompt-security/clawsec --skill clawsec-suite -a openclaw --global -y

This installs the suite with its signed advisory trust set, heartbeat workflow, guarded installer, and setup scripts. Optional protections remain separate packages that the suite discovers from the published catalog.

2. Review and enable the advisory hook

SUITE_DIR="${INSTALL_ROOT:-$HOME/.openclaw/skills}/clawsec-suite"
node "$SUITE_DIR/scripts/setup_advisory_hook.mjs"

The setup script prints its preflight before it changes persistent OpenClaw configuration. After it succeeds, restart the OpenClaw gateway and run `/new` once to trigger the first advisory scan.

To see the current optional protections:

node "$SUITE_DIR/scripts/discover_skill_catalog.mjs"

> **Installing for someone else?** Ask their agent to install `clawsec-suite` with the command above, show the hook preflight, and wait for approval before enabling the hook or optional cron job.

<details> <summary><strong>Shell and path notes</strong></summary>

For `bash` and `zsh`, keep home variables expandable:

export INSTALL_ROOT="$HOME/.openclaw/skills"

Do not single-quote paths that contain `$HOME`. In PowerShell, build the path explicitly:

$env:INSTALL_ROOT = Join-Path $HOME ".openclaw\skills"
node "$env:INSTALL_ROOT\clawsec-suite\scripts\setup_advisory_hook.mjs"

POSIX `.sh` workflows require WSL or Git Bash on Windows.

</details>

---

What was scanned

The files the safety scan read, as written by the repository's author and shown as plain text. Lines with a finding are marked.

.claude-plugin/marketplace.json · Manifest · 56 lines

View this file on GitHub (read at the repository's current version: the scan did not record which commit it read)

This file has changed since it was scanned, so this is not exactly the text that was scanned, and no line is marked. The findings are listed with the safety scan above.

  1. {
  2. "plugins": [
  3. {
  4. "name": "global-skills",
  5. "source": "./",
  6. "skills": [
  7. "./skills/clawtributor"
  8. ]
  9. },
  10. {
  11. "name": "hermes-skills",
  12. "source": "./",
  13. "skills": [
  14. "./skills/hermes-attestation-guardian",
  15. "./skills/hermes-traffic-guardian"
  16. ]
  17. },
  18. {
  19. "name": "nano-claw-skills",
  20. "source": "./",
  21. "skills": [
  22. "./skills/clawsec-nanoclaw",
  23. "./skills/nanoclaw-traffic-guardian"
  24. ]
  25. },
  26. {
  27. "name": "open-claw-skills",
  28. "source": "./",
  29. "skills": [
  30. "./skills/clawsec-clawhub-checker",
  31. "./skills/clawsec-feed",
  32. "./skills/clawsec-scanner",
  33. "./skills/clawsec-suite",
  34. "./skills/openclaw-audit-watchdog",
  35. "./skills/openclaw-traffic-guardian",
  36. "./skills/soul-guardian"
  37. ]
  38. },
  39. {
  40. "name": "pico-claw-skills",
  41. "source": "./",
  42. "skills": [
  43. "./skills/picoclaw-security-guardian",
  44. "./skills/picoclaw-self-pen-testing",
  45. "./skills/picoclaw-traffic-guardian"
  46. ]
  47. },
  48. {
  49. "name": "repo-internal-skills",
  50. "source": "./",
  51. "skills": [
  52. "./skills/claw-release"
  53. ]
  54. }
  55. ]
  56. }
README.md · README · 205 lines

View this file on GitHub (read at the repository's current version: the scan did not record which commit it read)

This file has changed since it was scanned, so this is not exactly the text that was scanned, and no line is marked. The findings are listed with the safety scan above.

  1. <p align="center">
  2. <img src="./assets/readme/hero.webp" width="100%" alt="ClawSec security skills for AI agents, featuring the ClawSec robot and the Prompt Security from SentinelOne logo">
  3. </p>
  4. <p align="center">
  5. <a href="https://clawsec.prompt.security"><strong>Website</strong></a>
  6. ·
  7. <a href="https://clawsec.prompt.security/#/skills"><strong>Skill catalog</strong></a>
  8. ·
  9. <a href="https://clawsec.prompt.security/feed"><strong>Security feed</strong></a>
  10. ·
  11. <a href="./wiki/INDEX.md"><strong>Documentation</strong></a>
  12. ·
  13. <a href="https://github.com/prompt-security/clawsec/releases"><strong>Releases</strong></a>
  14. </p>
  15. <p align="center">
  16. <a href="https://github.com/prompt-security/clawsec/actions/workflows/ci.yml"><img src="https://github.com/prompt-security/clawsec/actions/workflows/ci.yml/badge.svg" alt="CI status"></a>
  17. <a href="https://github.com/prompt-security/clawsec/actions/workflows/deploy-pages.yml"><img src="https://github.com/prompt-security/clawsec/actions/workflows/deploy-pages.yml/badge.svg" alt="Pages deployment status"></a>
  18. <a href="https://github.com/prompt-security/clawsec/actions/workflows/poll-nvd-cves.yml"><img src="https://github.com/prompt-security/clawsec/actions/workflows/poll-nvd-cves.yml/badge.svg" alt="NVD polling status"></a>
  19. </p>
  20. ClawSec is an AGPL collection of security skills and signed advisory intelligence for AI agent runtimes. It helps operators verify skill artifacts, detect configuration drift, audit agent environments, and approval-gate risky installs across **OpenClaw, NanoClaw, Hermes, and Picoclaw**.
  21. ---
  22. ## Install the OpenClaw suite
  23. The OpenClaw entry point is `clawsec-suite`. Adding the package and activating its persistent hook are separate, reviewable steps.
  24. ### 1. Add the suite
  25. ```bash
  26. npx skills add prompt-security/clawsec --skill clawsec-suite -a openclaw --global -y
  27. ```
  28. This installs the suite with its signed advisory trust set, heartbeat workflow, guarded installer, and setup scripts. Optional protections remain separate packages that the suite discovers from the published catalog.
  29. ### 2. Review and enable the advisory hook
  30. ```bash
  31. SUITE_DIR="${INSTALL_ROOT:-$HOME/.openclaw/skills}/clawsec-suite"
  32. node "$SUITE_DIR/scripts/setup_advisory_hook.mjs"
  33. ```
  34. The setup script prints its preflight before it changes persistent OpenClaw configuration. After it succeeds, restart the OpenClaw gateway and run `/new` once to trigger the first advisory scan.
  35. To see the current optional protections:
  36. ```bash
  37. node "$SUITE_DIR/scripts/discover_skill_catalog.mjs"
  38. ```
  39. > **Installing for someone else?** Ask their agent to install `clawsec-suite` with the command above, show the hook preflight, and wait for approval before enabling the hook or optional cron job.
  40. <details>
  41. <summary><strong>Shell and path notes</strong></summary>
  42. For `bash` and `zsh`, keep home variables expandable:
  43. ```bash
  44. export INSTALL_ROOT="$HOME/.openclaw/skills"
  45. ```
  46. Do not single-quote paths that contain `$HOME`. In PowerShell, build the path explicitly:
  47. ```powershell
  48. $env:INSTALL_ROOT = Join-Path $HOME ".openclaw\skills"
  49. node "$env:INSTALL_ROOT\clawsec-suite\scripts\setup_advisory_hook.mjs"
  50. ```
  51. POSIX `.sh` workflows require WSL or Git Bash on Windows.
  52. </details>
  53. ---
  54. ## See it work
  55. ### Detect and respond to agent-file drift
  56. The `soul-guardian` demo changes a protected agent file, detects the mismatch, and walks through the response.
  57. [![ClawSec soul-guardian drift detection demo](public/video/soul-guardian-demo-preview.gif)](public/video/soul-guardian-demo.mp4)
  58. **[Watch the MP4 with audio →](public/video/soul-guardian-demo.mp4)**
  59. <details>
  60. <summary><strong>Watch the suite installation walkthrough</strong></summary>
  61. <p align="center">
  62. <a href="./public/video/install-demo.mp4"><img src="./public/video/install-demo-preview.gif" width="320" alt="ClawSec suite installation walkthrough"></a>
  63. </p>
  64. <p align="center"><a href="./public/video/install-demo.mp4"><strong>Open the MP4 with audio →</strong></a></p>
  65. </details>
  66. ---
  67. ## How ClawSec protects an agent
  68. | Protection layer | What it does |
  69. | --- | --- |
  70. | **Signed intelligence** | Verifies the advisory feed and checksum manifest before matching published risk against installed skills. |
  71. | **Guarded installs** | Stops on advisory matches and requires a second, explicit confirmation before a risky install can continue. |
  72. | **Integrity and drift** | Gives platform-specific skills baselines for critical files, configuration, attestations, and release artifacts. |
  73. | **Audits and reporting** | Provides focused audit, posture, self-test, and community-reporting packages where the platform contract supports them. |
  74. ClawSec recommends and gates actions; destructive removal and install overrides remain approval-controlled.
  75. ### Platform entry points
  76. - **OpenClaw** — start with [`clawsec-suite`](skills/clawsec-suite/) for signed advisory monitoring and guarded installs, then discover separate drift and audit protections.
  77. - **NanoClaw** — use [`clawsec-nanoclaw`](skills/clawsec-nanoclaw/) for NanoClaw-specific advisory, integrity, verification, and security-tool workflows.
  78. - **Hermes** — use [`hermes-attestation-guardian`](skills/hermes-attestation-guardian/) for signed advisory checks, guarded verification, deterministic attestations, and baseline drift detection.
  79. - **Picoclaw** — use [`picoclaw-security-guardian`](skills/picoclaw-security-guardian/) for posture, advisory, drift, and release-artifact checks. [Self-pen testing](skills/picoclaw-self-pen-testing/) is a separate opt-in package.
  80. > The `*-traffic-guardian` directories are specification baselines for platform builders. They are not shipped runtime proxies today.
  81. Browse every package in the **[live skill catalog](https://clawsec.prompt.security/#/skills)** or the repository’s **[`skills/` directory](skills/)**.
  82. ### Skill feature matrix
  83. The full package comparison is preserved in the wiki, including shipped, limited, and specification-only coverage.
  84. **[Compare every skill in the feature matrix →](wiki/skill-feature-matrix.md)**
  85. ---
  86. ## Query the signed advisory channel
  87. The consolidated feed can contain relevant NVD CVEs, approved community reports, and provisional GitHub advisories that do not yet have CVE identifiers. Daily NVD and GitHub advisory polling covers OpenClaw, NanoClaw, Hermes, Picoclaw, NVIDIA NemoClaw, and NVIDIA OpenShell. NVIDIA OpenShell is matched by its NVD CPE (`cpe:2.3:a:nvidia:openshell`), the `NVIDIA/OpenShell` repository, and the phrase "NVIDIA OpenShell". A bare OpenShell mention is not enough, so OpenClaw's OpenShell component and unrelated projects that use the same word stay in their own buckets.
  88. ```bash
  89. curl -fsSL https://clawsec.prompt.security/advisories/feed.json \
  90. | jq '.advisories[] | select(.severity == "critical" or .severity == "high")'
  91. ```
  92. Trust material lives beside the feed:
  93. - [Advisory feed](advisories/feed.json)
  94. - [Detached feed signature](advisories/feed.json.sig)
  95. - [Pinned Ed25519 public key](advisories/feed-signing-public.pem)
  96. - [Signing and verification runbook](wiki/security-signing-runbook.md)
  97. The legacy `/releases/latest/download/feed.json` endpoint remains a compatibility mirror. New consumers should use the canonical `/advisories/feed.json` endpoint.
  98. ---
  99. ## Build, test, and contribute
  100. Run the web catalog locally:
  101. ```bash
  102. npm install
  103. npm run dev
  104. ```
  105. Run the repository’s local quality gate before pushing:
  106. ```bash
  107. ./scripts/prepare-to-push.sh
  108. ```
  109. Validate a skill package directly:
  110. ```bash
  111. python utils/validate_skill.py skills/clawsec-feed
  112. ```
  113. Start with these references:
  114. - [Architecture](wiki/architecture.md)
  115. - [Platform verification](wiki/platform-verification.md)
  116. - [Testing](wiki/testing.md)
  117. - [Release automation](wiki/modules/automation-release.md)
  118. - [Contributing guide](CONTRIBUTING.md)
  119. - [Security policy](SECURITY.md)
  120. The source of truth for project documentation is [`wiki/`](wiki/). GitHub Wiki pages and LLM-ready exports are generated from those files.
  121. ---
  122. ## Translations
  123. **English**
  124. · [Deutsch](README.de.md)
  125. · [Español](README.es.md)
  126. · [Français](README.fr.md)
  127. · [日本語](README.ja.md)
  128. · [한국어](README.ko.md)
  129. Localized wiki indexes: [DE](wiki/de/INDEX.md) · [ES](wiki/es/INDEX.md) · [FR](wiki/fr/INDEX.md) · [JA](wiki/ja/INDEX.md) · [KO](wiki/ko/INDEX.md) · [EN](wiki/INDEX.md)
  130. ---
  131. ## License
  132. ClawSec source code is licensed under **GNU AGPL-3.0-or-later**. See [LICENSE](LICENSE). Files under [`font/`](font/) have separate license terms and are not used by the README artwork.
  133. <p align="center">
  134. <strong>ClawSec</strong> · Prompt Security, from SentinelOne<br>
  135. Verify before your agent trusts.
  136. </p>

Ratings and comments

Sign in to rate this skill.

Sign in

Free to use. If it’s helped, pass it on.