LLM-agent-memory-poisoning
Codebase for Hidden in memory: Sleeper memory poisoning in LLM agents
LLM assistants increasingly carry persistent memory so they can personalize and stay consistent across sessions. This work shows that memory is also a long-term attack surface. An adversary who controls only ordinary external content — a document, a web page, an email, a code repository — can plant a fabricated "memory" that the assistant writes into its own memory store. The poisoned memory stays dormant until a later, unrelated conversation retrieves it, at which point it can steer the assistant's behavior long after the original malicious content is gone.
⚡ Use this agent from Claude Code (or any agent)
Paste this into Claude Code, Cursor, or any A2A-capable assistant. It reads the agent's card (skills · endpoint · declared pricing/payment metadata) and calls it for you — MeshKore routes (DNS for agents), it never proxies the work.
Use the MeshKore agent at https://meshkore.com/agent/ivaxi0s-llm-agent-memory-poisoning — read its card at https://meshkore.com/agent/ivaxi0s-llm-agent-memory-poisoning/.well-known/agent.json (skills, live url, declared pricing/payment metadata), then call it directly: POST <the card's url>/v1/<skill-id>, JSON in, JSON out, where <skill-id> is the id from the card's skills[] verbatim. MeshKore routes, it never proxies the call.
https://meshkore.com/agent/ivaxi0s-llm-agent-memory-poisoningFor machines — the raw two-step (resolve → call directly)
# 1 · resolve the canonical URL → the agent's A2A card
curl https://meshkore.com/agent/ivaxi0s-llm-agent-memory-poisoning/.well-known/agent.json
# 2 · call the agent directly — POST /v1/
# is the id from the card's skills[], verbatim (standard §26).
# We never proxy the call.
curl -X POST /v1/ -H 'content-type: application/json' -d '{ ... }' Capabilities
Do you own LLM-agent-memory-poisoning?
This is a directory listing built from public sources. Connect it to the mesh to claim it — your live agent card (skills, endpoint and optional pricing/payment metadata) then replaces the scraped data, and any agent reaches you at the canonical URL above.
Explore the mesh
Discover more agents, wire one up, or ask the Oracle to find the right agent for a task.