369 tagged AI Safety, measured the same way as everything else here.
Browse within: ai-security 267ai-behavior-analysis 149ai-safety-research 149ai-skill-safety 148ai-governance 139agent-governance 125policy-as-code 107credentials 106Guardrails 77claude-code-plugin 68ai-tools 64ai-coding 63compliance 59presentation 56
Skill Claude CodeCodex needs its repo
7-layer AI security + ML detection for OpenClaw. Covers all 10 OWASP Agentic AI risks — prompt injection, tool misuse, memory poisoning, data exfiltration, and more — across ALL channels (Signal, Telegram, Discord, WhatsApp, web) simultaneously.
Skill Claude Code
Write a warrant-mcp policy in plain English, shaped for the closed rule set so review has the best chance of accepting it first time — only warrant-mcp review can say whether it compiles. Use when the user wants to create or edit .warrant/policy.md, write rules for what an AI agent may do in a project, or asks what…
Skill Claude CodeCodex
Gate high-stakes, factual, or multi-step Hermes tasks before action and verify completion against evidence.
Skill Claude CodeCodex
Use when a user wants to complete METTLE reverse-CAPTCHA challenges, obtain a signed result, or verify a METTLE credential.
Skill Claude CodeCodex
Use when creating a blog post from scratch. Guides the full process from topic research through to a publish-ready draft with proper SEO structure and heading hierarchy.
Saenai/microsoft-todo-safe-mcp
Skill Codex
Inspect, validate, operate, and develop a self-contained Microsoft To Do Safe MCP skill with bundled server. Use for Microsoft To Do task/list management, safe backup/plan/preview/apply workflows, MCP startup checks, token-safe diagnostics, and deciding whether to run doctor/test/typecheck/build.
Skill Claude CodeCodex
Safely inspect a repository through the ContextLock MCP server. Use when exploring, searching, reviewing, or auditing repository files with ContextLock, especially before reading unfamiliar code or when sensitive files and secrets may be present.
joy7758/titmas-agent-action-gate
Skill Claude CodeCodex
Normalize an ambiguous external-action proposal into a versioned request while preserving uncertainty and granting no authority.
At most 3 mods per repository are shown here, and a mod shipped inside a plugin is left to that plugin's page — the rest are on their repository pages: