Ori Memory
A knowledge graph implementation that treats agent memory as a navigation problem rather than a search problem. Exposes tools for graph traversal with PageRank, semantic embedding fusion, and ACT-R activation decay. Notes live as markdown files with wiki-links as edges, segmented into identity, knowledge, and operations with different decay rates. Includes recursive retrieval that decomposes complex queries into sub-questions and learns from usage patterns through Hebbian co-occurrence. The system prunes intelligently using Tarjan's algorithm to protect structurally critical nodes. Reach for this when you need persistent memory that survives across sessions and clients without external infrastructure, especially for multi-hop reasoning tasks where context needs to propagate through connected concepts rather than flat similarity scoring.
- tools
- 15
- transport
- STDIO
- last updated
- Feb 27, 2026
Tools
15 toolsPublic tool metadata for what this MCP can expose to an agent.
ori_wakeSession boot: bounded briefing, plus onboarding on a fresh vault
Session boot: bounded briefing, plus onboarding on a fresh vault
No parameter schema in public metadata yet.
ori_updateWrite to identity, goals, methodology, daily, or reminders
Write to identity, goals, methodology, daily, or reminders
No parameter schema in public metadata yet.
ori_update_decisionRecord the user's answer to an update notice
Record the user's answer to an update notice
No parameter schema in public metadata yet.
ori_whats_newRelease notes for the installed version, or any version
Release notes for the installed version, or any version
No parameter schema in public metadata yet.
memory_sqlRead-only SQL over the index — anything the ranking tools cannot express
Read-only SQL over the index — anything the ranking tools cannot express
No parameter schema in public metadata yet.
ori_healthFull diagnostics
Full diagnostics
No parameter schema in public metadata yet.
ori_addCapture to inbox
Capture to inbox
No parameter schema in public metadata yet.
ori_promotePromote with classification, linking, and area assignment
Promote with classification, linking, and area assignment
No parameter schema in public metadata yet.
ori_validateSchema validation
Schema validation
No parameter schema in public metadata yet.
ori_query_rankedFull retrieval with Q-value reranking, co-occurrence PPR, and stage meta-learning
Full retrieval with Q-value reranking, co-occurrence PPR, and stage meta-learning
No parameter schema in public metadata yet.
ori_query_similarSemantic search (vector only, faster)
Semantic search (vector only, faster)
No parameter schema in public metadata yet.
ori_warmthInspect the associative warmth field
Inspect the associative warmth field
No parameter schema in public metadata yet.
ori_exploreRecursive graph traversal — PPR, sub-question decomposition, convergence detection
Recursive graph traversal — PPR, sub-question decomposition, convergence detection
No parameter schema in public metadata yet.
ori_pruneActivation topology analysis and archive candidates
Activation topology analysis and archive candidates
No parameter schema in public metadata yet.
ori_index_buildBuild/update embedding index and bootstrap co-occurrence edges
Build/update embedding index and bootstrap co-occurrence edges
No parameter schema in public metadata yet.
Sponsors
Agent CaffeineSponsoredClose your lid, come back to finished work
SnugSponsoredThe clipboard manager for the AI eraPromote your product hereReach 125k+ visitors monthly
WaxSponsoredBlock anything with supervision, without erasing your iPhonePromote your product hereReach 125k+ visitors monthly
knobsSponsoredSee your site on any phone, any day, in any languageReadme
Ori Mnemos
Open-source persistent memory infrastructure for AI agents.
Ori implements human cognition as mathematical models on a knowledge graph. Activation decay from ACT-R. Spreading activation along wiki-link edges. Hebbian co-occurrence from retrieval patterns. Reinforcement learning on retrieval itself. Recursive graph traversal with sub-question decomposition. The system learns what matters, forgets what doesn't, and optimizes its own retrieval pipeline.
Persistent memory across sessions, clients, and machines. Zero-infrastructure retrieval that matches and in several cases strongly outperforms incumbents on benchmarks — and you own every byte of your data. Markdown on disk. Wiki-links as graph edges. Git as version control. No database lock-in, no cloud dependency, no vendor capture.
v0.7.0 · npm · Paper · Apache-2.0
Use
Ori is three surfaces over one index. The markdown is the truth; the index is derived. The learned half — Q-values, LinUCB arms, retrieval history — is not, so export it before deleting anything (see When to rebuild).
Details
- Repository
- github.com/aayoawoyemi/ori-mnemos
- Categories
- AI & LLM Tools, Cloud & Infrastructure, Documents & Knowledge, Search & Web Crawling
- Package
ori-memory- Transport
- STDIO
- Registry
- MCP registry, active
Related AI & LLM Tools MCP servers
View all
3D AI Agent Avatarnirholas/three.ws3D AI Agent Avatar — render any GLB, give it a Solana wallet, a voice, and pump.fun powers.269stars
Neural Memorynhadaututtheky/neural-memoryPersistent memory for AI agents — 55 MCP tools, spreading activation recall.202stars
OMEGA Memoryomega-memory/omega-memoryPersistent memory, coordination, and learning for AI agents. Local-first, 25 MCP tools.158stars
MCP Memory Gatewayigorganapolsky/mcp-memory-gatewayPre-action gates that block AI agents from repeating known mistakes.23stars
Infomaniak Mcp Agentmogacode-ma/infomaniak-mcp-agentUnofficial Infomaniak MCP — hosting, Node.js, mail, DNS, kDrive, SSL, VPS, domains. 81 tools.8stars
Anneal Memoryphillipclapham/anneal-memoryTwo-layer memory for AI agents with an immune system. Zero dependencies.8stars