Now liveThe Skillselion MCP - thousands of ranked skills, loaded into your agent mid-task. No install.Get it →

prompt-security/clawsec

16 skills3.3k installs17.3k starsGitHub

Install

npx skills add https://github.com/prompt-security/clawsec

Skills in this repo

1Clawsec SuiteClawSec Suite is a ClawSec suite manager skill (version 0.1.8) that hardens OpenClaw-based agent workflows against untrusted skills. Setup installs an advisory hook under ~/.openclaw/hooks, optionally schedules an openclaw cron job, and can pull guard skills via npx clawhub@latest install. Runtime requires node, npx, openclaw, curl, jq, shasum, openssl, and unzip. Developers reach for ClawSec Suite when onboarding community skills into Claude Code or Cursor pipelines and need signature-checked, approval-gated blocking instead of running unknown SKILL.md payloads blindly.761installs2Openclaw Audit Watchdogopenclaw-audit-watchdog version 0.1.9 from prompt-security/clawsec automates daily security audits for OpenClaw agents. On invocation it creates or updates an `openclaw cron` job that runs `openclaw security audit --json` and `openclaw security audit --deep --json`, summarizes critical, warning, and info findings, and delivers formatted reports via configured DM channel and recipient plus optional email. Default schedule is daily at 23:00 in a chosen IANA timezone. Required environment variables are PROMPTSEC_DM_CHANNEL and PROMPTSEC_DM_TO; optional PROMPTSEC_EMAIL_TO enables email copies. Runtime requires bash, openclaw, and node. Standalone installs support signed release verification against checksums.json, checksums.sig, and signing-public.pem before extraction. Developers reach for openclaw-audit-watchdog when OpenClaw deployments need recurring unattended security posture checks with ClawSec policy visibility.686installs3Soul GuardianSoul-guardian from prompt-security/clawsec helps teams secure agent personas and system prompts against injection, override, and identity drift. It guides defensive prompt design, soul-boundary enforcement, and hardening patterns for production LLM agents.244installs4Clawsec Clawhub CheckerClawsec-clawhub-checker from prompt-security/clawsec provides automated security review for ClawHub agent skills, flagging risky prompts, excessive permissions, and policy issues so teams can safely ship extensions and CLI-distributed tooling through formal ship-phase security checks.232installs5Clawsec Feedclawsec-feed connects prompt-security ClawSec repositories to ongoing advisory monitoring for AI and API products. It helps teams track prompt-injection trends, CVEs, and security bulletins so agent prompts, tool policies, and deployments can be updated before incidents spread across production environments.231installs6ClawtributorSupports Clawsec contributor governance by systematically evaluating inbound dependencies and contributions: verify licenses and attribution obligations, map known CVEs and fix paths, assess maintainer reputation and update cadence, and record accept, patch, or reject decisions so security teams maintain a defensible supply-chain audit trail across agent tooling releases.197installs7Claw ReleaseCoordinates secure release for Clawsec-aligned projects: define mandatory security checks before tag or publish, scan dependencies and configs for known risks, verify secrets are absent from artifacts, document exceptions, and produce an auditable release record so agent skills and security tooling ship with consistent appsec standards rather than ad-hoc manual review.179installs8Clawsec Scannerclawsec-scanner from prompt-security/clawsec runs automated security audits on prompts, skills, and agent configurations, flagging injection paths, exfiltration patterns, and policy breaches before release.179installs9Clawsec Nanoclawclawsec-nanoclaw delivers a compact security layer from prompt-security/clawsec for agent and skill integrations, enforcing prompt-injection and unsafe-tool-use guardrails with low overhead during build.177installs10Hermes Attestation Guardianhermes-attestation-guardian is a Claude Code skill for testing & qa. It helps solo builders move faster with AI-assisted development.90installs11Picoclaw Security Guardianpicoclaw-security-guardian is a Claude Code skill for security. It helps solo builders move faster with AI-assisted development.54installs12Picoclaw Self Pen Testingpicoclaw-self-pen-testing is a Claude Code skill for testing & qa. It helps solo builders move faster with AI-assisted development.54installs13Hermes Traffic Guardianhermes-traffic-guardian is a Claude Code skill for ai & agent building. It helps solo builders move faster with AI-assisted development.48installs14Openclaw Traffic Guardianopenclaw-traffic-guardian is a Claude Code skill for ai & agent building. It helps solo builders move faster with AI-assisted development.41installs15Nanoclaw Traffic Guardiannanoclaw-traffic-guardian is a Claude Code skill for ai & agent building. It helps solo builders move faster with AI-assisted development.39installs16Picoclaw Traffic Guardianpicoclaw-traffic-guardian is a Claude Code skill for ai & agent building. It helps solo builders move faster with AI-assisted development.39installs

This week in AI coding

Five minutes, every Monday - the tools, releases and tactics for developers.

unsubscribe anytime.