Library
85,762 resources from 11 public sources. 54 checked by hand; the rest are unreviewed.
Results
-
Lycheetah-Framework
SkillsPublic method documents. How the work is thought and governed. Self-taught, in Dunedin. Not a commercial offer.
git clone https://github.com/Lycheetah/Lycheetah-Framework.git -
lintlang
PromptsStatic analysis for AI agent configs, tool descriptions, and system prompts — catches vague tool descriptions, missing stop conditions, and schema gaps before they reach runtime. Zero-LLM, deterministic checks, built for CI.
git clone https://github.com/hermes-labs-ai/lintlang.git -
phantom-secrets
MCP serversStop AI coding agents from leaking your API keys. Local proxy + MCP that swaps real secrets for phm_ tokens — works with Claude Code, Cursor, Windsurf, and Codex.
git clone https://github.com/ashlrai/phantom-secrets.git -
patton-protocol
SkillsAgent Skills package for bounded coding-agent missions, and a Claude Code hook that asks before existing files change. Pre-registered study: the hook cut cleanup harm from 0.196 to 0 over 224 runs; the skill text alone did not.
git clone https://github.com/vladimirrott/patton-protocol.git -
verbatim-citation-gate
SkillsCatch fabricated RAG citations before they reach the user: a zero-token verbatim gate + a burden-of-proof judge, framework-agnostic.
git clone https://github.com/tonydzi/verbatim-citation-gate.git -
creedspace-mcp-server
MCP serversConstitutional-AI safety guardrails for any LLM — personas, constitutions, adjudication.
@creedspace/mcp-server@1.1.4 -
humane-proxy
MCP serversAI safety middleware — detects self-harm and criminal intent in LLM prompts.
humane-proxy@0.6.0 -
agent-leash
SkillsLEASH-8: an 8-domain control model for AI agents with delegated authority. Scorecard, approval-design checklist, plan-vs-authorize pattern. Patterns from a live production agent operation, sanitized. Free, MIT.
git clone https://github.com/tonydzi/agent-leash.git -
ai-watermark-scrambler
SkillsScramble statistical AI text watermarks by rewriting with a local model (vLLM, Ollama), and strip C2PA, EXIF, XMP and invisible Unicode. Agent skill plus Python CLI.
git clone https://github.com/novatechflow/ai-watermark-scrambler.git -
postgres-scout-mcp
MCP serversScout your PostgreSQL databases with AI - safety features, monitoring, and data quality
postgres-scout-mcp@1.0.3 -
mongo-scout-mcp
MCP serversScout your MongoDB databases with AI - safety features, live monitoring, and data quality
mongo-scout-mcp@1.0.3 -
agentguard
MCP serversAgentGuard — 20-tool AI safety MCP: policy preflight, risk scoring, audit logging, rate limits.
-
healthguard
MCP serversHealthGuard - 12-tool health/medical AI safety MCP: PII redaction, HIPAA, GDPR Art.9.
-
oasis
SkillsThe open-source coding agent for any model: terminal and self-hosted web app, Claude Code plugins, subagents, and work that is actually verified.
git clone https://github.com/FAHTHIZ/oasis.git -
AAAA-Nexus MCP
MCP serversFormally verified AI safety APIs. 75+ endpoints, pay-per-call via USDC x402, no signup.
aaaa-nexus-mcp@0.1.1 -
mcp-regulated-ai-compliance
MCP serversRegulated-industry AI compliance: EU AI Act, APRA, NIST AI RMF, ISO 42001, AU AI Safety.
@hellouchit/mcp-regulated-ai-compliance@0.2.3 -
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 -
aizee
SkillsThe policy layer for AI coding agents. aiZee governs what Cursor, Claude Code, Copilot, Windsurf, Devin & MCP agents may do — 139 enforcement gates, 29 personas, 245 skills, BMAD runtime, hybrid memory, tamper-evident audit. pip install aizee
git clone https://github.com/m3taz-ahmed/aizee.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 -
AI/LLM agent output audit MCP: policy eval, tamper-evident chain, AI safety, x402 USDC on Base.
-
ai-security-corpus
Skills5,660 AI security links: prompt injection, LLM & AI agent security, MCP, coding agents, agent skills & supply chain, red teaming, guardrails, sandboxing, incidents & CVEs, governance. 570 arXiv papers, 1,155 GitHub repos, 570 X posts, 29 categories of plain markdown you can grep, diff and fork. Obsidian vault export. Added, not endorsed.
git clone https://github.com/AnthonyHerman/ai-security-corpus.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.