Library
85,340 resources from 11 public sources. 54 checked by hand; the rest are unreviewed.
Results
-
agent-flow
SkillsGuardrails for AI coding agents: run Claude Code, Codex, Gemini CLI, Cursor or Copilot unattended and come back to a reviewed pull request. Implement, review and QA agents, a guard on protected paths, and a tamper-evident audit log.
git clone https://github.com/Drix10/agent-flow.git -
little-canary
SkillsDetects prompt injection by its effect on a sacrificial canary model, not just pattern matching: untrusted input hits a powerless model first, a behavioral check reads the residue, and it returns block, flag, or pass before your primary model acts. Inbound preflight sensor, not a guarantee.
git clone https://github.com/hermes-labs-ai/little-canary.git -
priors
SkillsOn-chain credit pool for ERC-8004 AI agents on Robinhood Chain: agents borrow USDG, repay it and earn a public credit score. Contracts, SDK, CLI, MCP server (@priors/mcp), x402 helpers (@priors/x402) and an agent skill.
git clone https://github.com/priors-agents/priors.git -
web3-agent-skills
Skills⚡ Drop-in SKILL.md, .cursorrules, CLAUDE.md, AGENTS.md & Remote MCP (.mcp.json) pack for Claude Code, Cursor, Windsurf, Cline & x402 AI Agents — Verify GitHub Bounties & Mine DOOM/wDOOM on Base Mainnet.
git clone https://github.com/doomsday-network/web3-agent-skills.git -
orrery
SkillsAnimated architecture and system design diagrams with official AWS, Azure, Google Cloud, Kubernetes, GitHub and IBM icons. YAML in; SVG, PNG, draw.io, React TSX and HTML out. An Agent Skill, Claude Code plugin and npm CLI. Gallery: sagargupta.online/orrery
git clone https://github.com/Sagargupta16/orrery.git -
contrarian
SkillsDevil's advocate for coding agents. Stress-tests plans, architecture and decisions with a steel-man, an assumption audit and a pre-mortem, then gives severity-rated findings and a verdict. Works in Claude Code, Codex, Cursor, Gemini CLI, Copilot and 12 more.
git clone https://github.com/aaddrick/contrarian.git -
skills
SkillsVideo skills for Claude Code, Codex, Grok, Cursor and Gemini CLI: 83 AI video skills for UGC ads, product videos, music videos, consistent characters, captions, motion graphics and FFmpeg edits. Seedance 2.5, Kling 4, Veo 3.1, GPT Image 2.6 and ElevenLabs on one key, or free and local.
git clone https://github.com/vargHQ/skills.git -
hermes-jailbench
SkillsZero-LLM deterministic jailbreak regression benchmark: runs a repeatable battery of known-pattern attacks against an LLM endpoint and scores refusal/partial/compliance across runs. Catch when a model or prompt update silently got weaker. Single-turn, responsible-use. pip install hermes-jailbench
git clone https://github.com/hermes-labs-ai/hermes-jailbench.git -
google-flow-api
SkillsNode.js and Python examples for the third-party Google Flow API by useapi.net: Veo 3.1, Omni 1.2 and Omni 1.1 Flash video with extend, Nano Banana Pro, 2.1 and 2 Lite images, Google Flow Agent chat (live demo app), n8n templates and a hosted MCP server, on your Google AI subscription. No Google Cloud project or Gemini API key.
git clone https://github.com/useapi/google-flow-api.git -
since-cutoff
SkillsFind which APIs of your pinned Python dependencies changed after your coding model's training cutoff, and give the agent short AGENTS.md / CLAUDE.md notes from a static API diff. No model calls, no API key. CLI, MCP server, Claude Code plugin, GitHub Action.
git clone https://github.com/MohammadHijjawi97/since-cutoff.git -
humanizer-it
SkillsRemove AI traces from Italian text — opensource (MIT) skill for Claude. The «IA-taliano»: English calques, nominal burocratese, missing intercalari/clitics. 52 patterns, 15 hard bans, dependency-free scanner. | Скилл для итальянского текста: убирает следы нейросети, 52 паттерна и 15 жёстких банов, сканер без зависимостей.
git clone https://github.com/ilyautov/humanizer-it.git -
autorun
SkillsClaude Code & codex plugin + Gemini cli extension to keep ai running, keep data safe, plan better, and keep running until tasks are done.
git clone https://github.com/ahundt/autorun.git
Indexed entries aren't reviewed or tested by InnerVault. Details come straight from each source. Anything that can reach your accounts, files or browser can do damage, so read the source before you install it.