Ratine
https://github.com/goweft/ratine
Only tool found that scans the persistent memory layer on disk rather than runtime tracing. Detects injected instructions, obfuscated payloads (zero-width Unicode, base64, homoglyphs, hex), C2-pattern URLs, credential leakage. ratine diff compares snapshots for belief drift.
Active · Verified 2026-10-22 · fresh as of 2026-10-22 · 43 cited facts, 0 with a third-party source · Suggest a correction
At a glance
- Tier
- T2
- Latest release
- v0.1.0 2026-04-02 repo fix
- Pricing
- free / open source (MIT); no paid tier repo fix
- Funding
- no public funding data found repo fix
- Freshness
- 2026-10-22 fix
- Layer
- agent-infra repo fix
- Licensing
- open-source repo fix
- Maintainer health
- 0 issues in 90 days · 1 release in 12 months (latest 2026-04-02) — as of 2026-11-01 repo
Memory taxonomy
- storage
- file
- retrieval
- extraction-pull
- persistence
- cross-session
- update
- read-only
- unit
- fact
- governance
- auditable
- conflict
- n/a
When to use
Optimised for: memory operation tracing + drift / poisoning detection
Anti-fit: not for use cases that don't run agent workloads in production
Pros & cons
Pros
Memory-specific observability with explicit memory-quality dashboards.
Cons
Newer entrant; small adoption signal; thin documentation.
Claims & capabilities
Free / open source. Auto-detects Mem0 local, AutoGen, custom layouts. Zero deps, zero network calls.
Technical surface
- API surface
- searched not found
- Backend storage
- searched not found
- Deployment
- local CLI — Python; pip install; no server required
- Embedding model
- searched not found
- Multi-tenancy
- searched not found
- MCP
- Yes — built for MCP tool-server ecosystem (sibling tools heddle/loom are MCP runtimes); detects agent memory poisoning across MCP sessions
- A2A
- no data — searched github.com/goweft/ratine README; single-purpose memory-poisoning detector for MCP tool servers, no A2A protocol
- OpenTelemetry
- searched not found
Alternatives to Ratine → live options ranked by shared design, licence, hosting and MCP support
Similar systems
Other memory observability & monitoring in the catalog, ranked by inbound references.
- Langfuse T1
Acquired by ClickHouse (2026-01-16); product and MIT-licensed repo continue under the Langfuse name. Memory module reads/writes captured as named spans. Trace Log View concatenates every agent step including memory ops. Agent Graphs (GA 2025) infer graph structure from observation nesting; session-level replay tracks how memory state evolves.
- LangSmith T1
Memory reads, vector DB retrievals, state changes are distinct span types in traces. RAG eval separates retrieval quality (context precision) from generation quality (faithfulness). Dataset versioning guards against eval drift.
- AgentOps T2
When Mem0 is connected, gains Memory Operation Timeline, Search Analytics, Memory Growth tracking, Error Tracking per memory call. Standalone, records context at each step but doesn't analyse memory quality.
- Galileo (galileo.ai) (now Splunk Agent Observability) T1
Acquired by Cisco (announced 2026-04-09, closed by 2026-05); now Splunk Agent Observability. Treats memory as first-class in multi-agent tracing. Luna-2 SLMs (3B / 8B) scan every interaction for intent drift and belief drift; 20+ checks at sub-200ms latency. Catches when agent A's view of the world splits from teammate B's. OpenTelemetry-compatible.
Sources (42 cited facts, 0 third-party)
Each fact links to the page it came from; the badge says who published it (how to read sources).
- What it is Offline scanner — memory poisoning detector repo
- What it does & what's distinct Only tool found that scans the persistent memory layer on disk rather than runtime traci… repo
- Claims / benefits Free / open source. Auto-detects Mem0 local, AutoGen, custom layouts. Zero deps, zero ne… repo
- Modalities text repo
- Created 2026-04 repo
- Latest release v0.1.0 2026-04-02 repo
- License MIT repo
- GitHub 1★ Python repo
- Funding no public funding data found repo
- Customers / scale not disclosed — vendor does not publish customer list repo
- Pricing free / open source (MIT); no paid tier repo
- Deployment local CLI — Python; pip install; no server required repo
- HQ Los Angeles, CA, US (developer Steve Gonzalez; per GitHub profile github.com/goweft) repo
- Founders / pedigree goweft (open-source; no institutional affiliation) repo
- Memory volume / scale not applicable (scanner; no memory storage) repo
- Performance no public benchmark scores found — vendor does not publish quantitative perf metrics repo
- Consistency strong repo
- Versioning none repo
- Schema evolution free-form / no schema repo
- Namespace primitives hierarchical repo
- MCP support Yes — built for MCP tool-server ecosystem (sibling tools heddle/loom are MCP runtimes); … repo
- Webhooks / events webhooks supported (alerting integrations) repo
- Import / export JSON via REST API repo
- Validated verticals agent developers + AI platform teams (horizontal) repo
- Time to running <10min (hosted SaaS) to <1hr (self-host) repo
- Anti-fit not for use cases that don't run agent workloads in production repo
- Optimised for memory operation tracing + drift / poisoning detection repo
- Adjacent infrastructure BYO agent stack; this layer observes it repo
- Pros Memory-specific observability with explicit memory-quality dashboards. repo
- Cons Newer entrant; small adoption signal; thin documentation. repo
- Project & sources GitHub goweft/ratine repo
- commit-trajectory [{"ym":"2026-04","cum":11}] repo
- cost-input-usd-per-mtok 0 repo
- cost-output-usd-per-mtok 0 repo
- cost-tier free repo
- cost-pricing-model free-tier repo
- cost-last-verified 2026-06-29 repo
- architecture-deployment-locus local repo
- licensing-bucket open-source repo
- positioning-platform-vs-product product repo
- taxonomy-three-layer-rollup agent-infra repo
- sdk-languages python registry
Related systems
References (1)
- Mem0 integrates with — Free / open source. Auto-detects Mem0 local, AutoGen, custom layouts.