Memory systems
OWASP Agent Memory Guard
Open-source runtime defense against memory poisoning
★ 204 on GitHub · last commit Nov 15, 2026
What it is
OWASP project targeting ASI-06 (Memory & Context Poisoning): scans memory reads and writes for injection payloads and structural tampering, keeps SHA-256 integrity baselines to catch post-write tampering, block/warn/strip policies; claimed sub-100 microsecond overhead. Also distributed as the agent-memory-guard pip middleware (discussed in microsoft/autogen#7792). GitHub API 2026-11-15: 204 stars, Apache-2.0, active (pushed 2026-11-15).
At a glance
- Business model
- Open source
- License
- Apache-2.0
GitHub activity
This product has a public repository, so we read it nightly from the GitHub API. Stars measure developer attention, not quality or market share.
“Runtime defense against memory poisoning in AI agents (OWASP ASI06). Drop-in for LangChain, OpenAI Agents, AutoGen, CrewAI and mem0.” (repository description)
- Stars
- 204
- Forks
- 64
- Open issues
- 27
- Last commit
- Nov 15, 2026
- License (GitHub)
- Apache-2.0
Links and sources
- Official siteowasp.org/projects/agent-memory-guard
- GitHub repositorygithub.com/OWASP/www-project-agent-memory-guard
- Sourcegithub.com/microsoft/autogen/discussions/7792
Last checked nov 15, 2026; last updated nov 15, 2026; facts verified nov 15, 2026. Vendor facts come from the vendor’s own pages; benchmark claims are self-reported unless we say otherwise.