570 tagged agent-security, measured the same way as everything else here.
Browse within: ai-security 325llm-security 255mcp-security 110formal-verification 80qualixar 79ai-agent-security 61agent-security-scanner 34Multi-Agent 33Prompt Injection 33ai-governance 33AI Safety 32agent-orchestration 32self-hosted 32owasp 31
Skill Claude CodeCodex
Read a documentation or API-reference page whose content is rendered by JavaScript (a single-page app) so that plain fetch/curl/WebFetch returns an empty shell, a loading spinner, or a 403 to script user-agents. Use to extract an API's surface — endpoints, methods, path/query/body params, schemas, auth scheme, base…
Skill Claude CodeCodex
Create new Comis skills, modify and improve existing skills, and validate them against the platform's manifest schema. Use this skill whenever the user wants to create a skill from scratch, turn a workflow into a reusable skill, update or fix an existing skill, understand the skill format, or asks about how skills…
Skill Claude CodeCodex
Observe and debug your OpenClaw agent with Korveo — local-first AI observability. Records every LLM call, tool call, and decision; view at localhost:3000. Use when the user asks: "show my agent traces", "why did my agent fail", "how much did my last run cost", "open korveo", "debug my agent", or "show recent traces".
Skill Claude CodeCodex
Integrate PancakeSwap swaps into applications. Use when user says "integrate swaps", "pancakeswap", "smart router", "add swap functionality", "build a swap frontend", "create a swap script", "smart contract swap integration", "use Universal Router", or mentions swapping tokens via PancakeSwap.
Skill Claude CodeCodex
Create a structured specification before agentic coding work. Assemble the six context types, scale rigor for prototype/internal/production tasks, produce SPEC.md, and configure focused AGENTS.md boundaries. Use when asked to write a spec, plan a feature, design an agent/system, define architecture, or create…
Skill Claude CodeCodex
Unified security suite for agent workspaces. Installs, configures, and orchestrates all 11 OpenClaw security tools in one command — integrity, secrets, permissions, network, audit trail, signing, supply chain, credentials, injection defense, compliance, and incident response.
Skill Claude CodeCodex
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.
doteyeso-ops/mcp-server-vibes-coded
Skill Claude CodeCodex
Use before installing an untrusted SKILL.md, MCP plugin, package script, or agent installer. Buys a deterministic supply-chain risk report for $0.05 USDC through FiatDock.
Skill Claude CodeCodex
Principles and checks for structuring code with clear boundaries, low coupling, and high cohesion. Use this whenever you are designing a new module, refactoring existing code, reviewing structure, deciding where logic should live, or untangling a messy codebase — even if the user doesn't say the words "clean…
Skill Claude CodeCodex
Reviewable synthetic skill that fetches a remote template before processing notes.
Skill Claude CodeCodex
Draft or repair a disabled Synapsor Safe Action when a developer asks to make one application data change safe for an AI agent.
Skill Claude CodeCodex
Use Nightward to audit local AI agent/devtool config, inspect MCP risk, and generate dry-run backup or schedule plans without leaking secrets.
Skill Claude CodeCodex
Multi-step workflow to audit MCPKernel documentation against the codebase and update README.md, docs/ MkDocs site, CHANGELOG.md. Use when synchronizing documentation with code changes, auditing doc accuracy, or before releases.
rudimentall1/agentic-wallet-guardian-v3
Skill Claude CodeCodex
Use before executing any MetaMask Agent Wallet (mm CLI) command that moves funds or changes on-chain state — mm send, mm swap, mm bridge, mm perps open/modify/close, mm predict trade, mm earn supply/withdraw, mm aave borrow/repay, mm pay. Sends the proposed action to a self-hosted Agentic Wallet Guardian instance for…
agentic-control-plane/codex-acp-plugin
Skill Claude CodeCodex
Generate a cost X-ray report for your agent from ACP's metering — cache economics, context composition, tool bloat, loop share — before you ship it (or whenever spend surprises you).
Skill Claude CodeCodex
Vet MCP tools/list and local AI-agent tool configuration before attachment. Use to check tool poisoning, forced invocation, confusable tool names, unconstrained shell or code inputs, risky fetch-plus-privilege combinations, delegation exposure, or elevated-permission requests with a pinned offline static gate.
Skill Claude CodeCodex
Vet a Model Context Protocol (MCP) server for safety before installing or connecting it. Use when the user wants to check, audit, scan, or review an MCP server for prompt-injection, data-exfiltration, or malicious capabilities — given a directory, npm package, GitHub repo, or a tools/list JSON dump.
professor2k8/agent-security-lens
Skill Claude CodeCodex
Requires BROWSERAPIKEY in .env.
Skill Claude CodeCodex
Review or harden an agent / MCP / local-first system using KASA's structural-defense methodology — mandatory threat-model actor tagging (A1-A4), deny-by-default / total-mediation checklist, measurement discipline (found->proved->closed, "detect+contain not prevent", no invented scores, positive+negative controls), and…
Skill Claude CodeCodex
Audit an AI agent deployment against the CUSTODY containment framework and LASM threat model. Checks identity, input handling, supervision, traceability, operational controls, dependency management, and yield/teardown. Maps findings to NIST AI RMF, OWASP, CUSTODY pillars, and LASM layers.
Skill Claude CodeCodex
Scan a public GitHub MCP-server repository for security issues before installing or connecting to it. Calls the hosted compuute-scan API at scan.compuute.se and returns severity counts, a 0-100 score, the most severe findings, and an honest triage disclaimer. Use BEFORE recommending or installing an unfamiliar MCP…
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: