OpenAgentSkill Registry Manifest
Skill: AgentPoison
Slug: ai-secure-agentpoison
Category: agent-frameworks
Description: [NeurIPS 2024] Official implementation for "AgentPoison: Red-teaming LLM Agents via Memory or Knowledge Base Backdoor Poisoning"

Agent fit:
- Decision: 80/100 Strong shortlist
- Primary fit: RAG and knowledge
- Role: Companion skill

Supply profile:
- Track: Coding and developer agents
- Scenario: GitHub automation
- Applicable agents: Claude Code, CLI, Codex, Cursor, Python
- Maintenance: 17d since push
- Risk: Safe to try

Trust:
- Trust score: 82/100 Strong shortlist
- Audit: 86/100 Safe to try

Attribution:
- Status: Community indexed
- Source: GitHub star discovery
- Creator: AI-secure
- Claim URL: https://www.openagentskill.com/skills/ai-secure-agentpoison#claim-this-skill

Install:
npx skills add AI-secure/AgentPoison

URLs:
- Web: https://www.openagentskill.com/skills/ai-secure-agentpoison
- API: https://www.openagentskill.com/api/agent/skills/ai-secure-agentpoison
- Install API: https://www.openagentskill.com/api/skills/ai-secure-agentpoison/install
- Repository: https://github.com/AI-secure/AgentPoison