Library
85,762 resources from 11 public sources. 53 checked by hand; the rest are unreviewed.
Results
-
Real-time MCP & A2A prompt-injection, jailbreak, and tool-schema poisoning firewall for AI agents.
-
Check a shop, checkout or payment link, paid API or token before paying; prompt-injection scan.
angel-agent-guard@0.2.0 -
agent-opfor
MCP serversOpen-source adversary emulation for AI agents and MCP servers.
git clone https://github.com/KeyValueSoftwareSystems/agent-opfor.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 -
Scan web pages, mails, PDFs and tool output for prompt injection before your agent reads them.
-
Learn-Prompt-Hacking
PromptsThis is The most comprehensive prompt hacking course available, which record our progress on a prompt engineering and prompt hacking course.
git clone https://github.com/TrustAI-laboratory/Learn-Prompt-Hacking.git -
Sponsio
SkillsDeterministic safety solutions for probabilistic AI agents
git clone https://github.com/SponsioLabs/Sponsio.git -
GenPark Prompt Injection Canary Leak Sentinel
MCP serversHMAC-signed zero-width invisible canary token generator and output response leak auditor
https://github.com/Alpha-Park/genpark-prompt-injection-canary-leak-sentinel-skill/releases/download/v1.0.1/genpark-prompt-injection-canary-leak-sentinel-skill.mcpb -
vestige
Agents & frameworksDeterministic Operating System and Causal Nervous System for Artificial Intelligence.
git clone https://github.com/samvallad33/vestige.git -
Prompt Injection Scanner
MCP serversCheck text before your agent reads it: hidden instructions, invisible text and links that leak data.
-
Habits prompt-injection guard
MCP serversCheck user prompts for injection/jailbreaks; your own guard learns from your labels. Works keyless.
-
Agent safety checks: prompt-injection scan, honeypot check, x402 preflight + crypto data (USDC/x402)
MCP serversChecks before an agent acts: prompt-injection scan, honeypot check, x402 preflight. 3 free calls.
-
agent-prompt-injection-firewall-mcp
MCP serversThe WAF for agents. Pattern-based + heuristic firewall scans prompts, RAG documents, tool argume...
agent-prompt-injection-firewall-mcp@1.0.13 -
nuguard
MCP serversAI red-teaming tool and LLM security framework to evaluate agentic AI applications. Tests prompt injections, handles vulnerability assessment, SBOM generation, and static analysis.
git clone https://github.com/NuGuardAI/nuguard.git -
invaris-agentsec
MCP serversAdversarial security testing for AI agents — prompt injection, tool misuse, memory poisoning, and MCP scanning, with CI-ready regression testing and reports.
git clone https://github.com/invarislabs/invaris-agentsec.git -
Awesome-Multimodal-Prompts
PromptsPrompts of GPT-4V & DALL-E3 to full utilize the multi-modal ability. GPT4V Prompts, DALL-E3 Prompts.
git clone https://github.com/langgptai/Awesome-Multimodal-Prompts.git -
35-probe LLM/agent security red-team scan (injection, jailbreak, MCP abuse) with report.
-
autoguardrails
Datasets & benchmarksAlignment-research scaffold (autoresearch-style) for LLM guardrails: search over a single policy.md surface
git clone https://github.com/SantanderAI/autoguardrails.git -
Awesome-GUI-Agent-Security
Agents & frameworksGUI / Computer-Use / Browser Agent 安全论文清单 —— 按攻防轴组织,每篇附中文简介
git clone https://github.com/Yuxuan2003/Awesome-GUI-Agent-Security.git -
mcp-scanner
MCP serversSecurity scanner for MCP servers, source trees, and packages, with explicit findings and coverage reporting.
git clone https://github.com/devlab-group/mcp-scanner.git -
llm-guard
PromptsThe Security Toolkit for LLM Interactions
git clone https://github.com/protectai/llm-guard.git -
sandlock
Agents & frameworksThe lightest sandbox. A process-based sandbox for Linux, no container, no VM, no privilege, no prompt injection
git clone https://github.com/multikernel/sandlock.git -
Hirepanel
Agents & frameworksMulti-agent resume screening where every score needs a verified quote from the resume. FastAPI, async DAG, n8n, Docker. Human approves every decision.
git clone https://github.com/punit2911/Hirepanel.git -
JailbreakHub
PromptsA Hub/research area, for jailbreaking LLM's, and breaking their safety filters.
git clone https://github.com/lalipa2003-arch/JailbreakHub.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.