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 Code
Discover the best Claude Code extensions on the web AND tell the user how safe each one is before installing. Reads the user's Claude Code setup (project context, memory, history), asks a few questions to understand their goal (general or a specific idea), then surfaces a ranked shortlist - name, what it does, what…
Skill Claude CodeCodex
This skill should be used when a request is ambiguous, underspecified, conflicting, or high impact. It is intended for vague verbs like optimize, improve, fix, refactor, and add feature; for missing file paths or unknown dependencies; and for risky actions like deploy, delete, overwrite, or migrate. It should not be…
Skill Claude CodeCodex
Detects whether a purchased AI API key/endpoint (Claude/Anthropic, OpenAI, etc.) — often bought cheap from a reseller or marketplace like GamsGo — is actually a man-in-the-middle proxy that intercepts, rewrites or degrades traffic. Trigger when the user says things like "I bought a third-party API key", "is this…
Skill Claude Code
Core agent-browser usage guide. Read this before running any agent-browser commands. Covers the snapshot-and-ref workflow, navigating pages, interacting with elements (click, fill, type, select), extracting text and data, taking screenshots, managing tabs, handling forms and auth, waiting for content, running multiple…
htlin222/research-guardian-skill
Skill Claude CodeCodex
A research-quality checking system for AI agents. It verifies facts, compares literature, checks experiment designs, and reviews claims about new findings.
levelsofself/mcp-nervous-system
Skill Claude CodeCodex
Use this skill when building, managing, or auditing multi-agent AI systems. Provides governance patterns for behavioral enforcement, drift detection, audit trails, role management, and accountability across autonomous AI agents. Compatible with any orchestration framework.
Skill Claude CodeCodex
A method for making an editable HTML slide presentation from an outline, notes, document, or verified facts. Each slide has a clear job, and the deck is designed for an audience and a set presentation length.
Skill Claude CodeCodex
Equilibrium-Native Reasoning for OpenClaw.
Skill Claude CodeCodex
Runtime guardrails for AI coding agents. Stop loops, budget overruns, retry storms, and timeouts before they burn money. Zero dependencies, local-first, MIT licensed.
Skill Claude CodeCodex
Improve and verify agent work after repeated failures, in dependency-heavy tasks, or when optimization claims need baseline and regression evidence. Use for DAGx/Perfectify requests, failed retries, risky multi-step work, or requests to verify an improvement. Exclude routine questions, drafting, one-step edits, and…
Skill Claude CodeCodex
A skill for using the local ClawHeart command-line tool to audit AI security, evaluate skills, and manage agent credentials.
Skill Claude Code needs its repo
AI agent guardrails — defends prompts against injection attacks, jailbreaks, and evasion; scans LLM responses for leaked secrets and harmful content. Up to 36 inline defenses (24 default), 3 output scanners, and adaptive ranking. Keywords: shield, guardrails, prompt injection, defense, security, scan.
Skill Claude CodeCodex
Generate draft text for one or more social platforms from a topic or URL. Wraps postall-agent generate. Activates when the user wish includes a content topic plus a publishing intent ("write a tweet about X", "draft a LinkedIn post on Y", "summarize this URL into a wechat article").
smq9sn5jck-coder/causallayer-mcp
Skill Claude CodeCodex
This skill provides instructions for AI agents interacting with the CausalLayer Model Context Protocol (MCP) server. CausalLayer is a deterministic AI liability apportionment engine. It takes an incident report and mathematically proves which agent, vendor, or operator is liable based on counterfactual do-calculus…
Skill Claude CodeCodex
Trust scoring, PII protection, and audit for OpenClaw agents. Strips personal data before it reaches the LLM. Blocks dangerous tools. Detects prompt injection. Audits everything.
zhengbingquant/frontier-skills
Skill Claude CodeCodex
Bootstrap a new Python project from a proven scaffold. Use whenever the user asks to start, create, scaffold, or set up a new Python tool, library, CLI, package, service, or experiment repo — or says 'set up a project the usual way'. Do NOT hand-roll a project structure from memory when this skill is available: run…
Skill Claude Code
Reviewable synthetic skill that fetches a remote template before processing notes.
levelsofself/palyan-agent-skills
Skill Claude CodeCodex
Use this skill when building, managing, or auditing multi-agent AI systems. Provides governance patterns for behavioral enforcement, drift detection, audit trails, role management, and accountability across autonomous AI agents. Compatible with any orchestration framework.
Skill Claude CodeCodex
Analyze repeated EvoPilot observations and compile proven workflows into portable, versioned Agent Skills or other reviewed extensions. Use when improving or extending the agent itself.
DennisWei9898/loop-engineering-reviewer
Skill Claude CodeCodex
A reviewer for automated agent or workflow projects, based on the Loop Engineering rules by Addy Osmani. It checks whether creation and review are separated, whether pass conditions are objective, and whether the loop has limits and human approval points.
razaumair2203-ux/codex-adversarial-review-lite
Skill Claude CodeCodex
Codex Adversarial Review - Lite: user-invoked audit workflow for Claude Code users who want Codex CLI to independently review AI-generated code, plans, test expectations, and scope before fixes are applied. Cross-platform (Windows, macOS, Linux, WSL). Use only when the user explicitly invokes audit or selftest.
Skill Claude CodeCodex needs its repo
A deliberately malicious demo Skill that tries to exfiltrate secrets over the network and read /.ssh/idrsa. Used to show Capsule blocking the calls at the call site.
gengshirong1128-boop/loop-governor
Skill Claude CodeCodex
A safety and decision layer for autonomous loops, where agents repeat work across multiple rounds. It sets limits, budgets, stop rules, file boundaries, and review points for workers such as Claude Code, Codex, and Gemini CLI.
marmar9615-cloud/agentbridge-protocol
Skill Claude CodeCodex
Operating guidance for an agent that has the AgentBridge MCP server available. The server exposes five tools, four resources, and four prompts. This skill explains when to use each tool and how to respect the safety contract.
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: