# Mnemo Cortex > Open-source cognitive coprocessor with active memory for AI agents β€” persistent recall, semantic search, overnight dreaming, verified facts, encrypted USB sync. MCP server; works with Claude, ChatGPT, and any local LLM. Built by one maker and his agents. - **Type:** MCP server - **Install:** `agentstack add mcp-guymanndude-mnemo-cortex` - **Verified:** Yes β€” security-reviewed for prompt injection and unsafe behavior - **Seller:** [GuyMannDude](https://agentstack.voostack.com/s/guymanndude) - **Installs:** 0 - **Category:** [Databases](https://agentstack.voostack.com/c/databases) - **Latest version:** 0.1.0 - **License:** MIT - **Upstream author:** [GuyMannDude](https://github.com/GuyMannDude) - **Source:** https://github.com/GuyMannDude/mnemo-cortex - **Website:** https://projectsparks.ai ## Install ```sh agentstack add mcp-guymanndude-mnemo-cortex ``` Requires the [AgentStack CLI](https://agentstack.voostack.com/docs/cli). Works with Claude Code, Cursor, and any MCP-compatible agent. ## About # ⚑ Mnemo Cortex β€” Cognitive Coprocessor with Active Memory [](https://glama.ai/mcp/servers/GuyMannDude/mnemo-cortex) **🌐 Home: [projectsparks.ai](https://projectsparks.ai)** β€” the story, the docs, and how the project is funded. > **The Project Sparks ecosystem** β€” everything is free; the whole thing runs on donations: > 🎨 [Free Art Gallery](https://rockysgallery.com) Β· πŸ§’ [Kids Seek-&-Find Gallery](https://rockysgallery.com/kids/) Β· 🎁 [Mnemo Swag](https://rockysgallery.com/swag/) Β· 🌐 [projectsparks.ai](https://projectsparks.ai) Β· ❀️ [PayPal](https://www.paypal.com/donate/?business=RHYJXYA2B877E&no_recurring=0&item_name=Support+Mnemo+Cortex+%E2%80%94+open-source+AI+memory%2C+funded+entirely+by+donations.¤cy_code=USD) / [GitHub Sponsors](https://github.com/sponsors/GuyMannDude) / [Ko-fi](https://ko-fi.com/guysparks) ## Why this exists The big AI-memory systems run on millions in venture capital. **Mnemo Cortex was built by one 73-year-old maker** β€” on Social Security, a $20-a-month Claude subscription, and a stubborn belief that his AI partners should remember him. It is not a demo. It is the production memory of a working five-agent fleet on two machines, every day since March 2026: - **~10,000 memories** ingested, compacted and recalled across agents - **~7,000 verified facts** with confidence laddering and a full audit history - **84 shipped versions** in five months ([CHANGELOG](CHANGELOG.md)) β€” currently v4.14.0 - Overnight dreaming, USB courier sync with encryption, agent-to-agent sharing β€” each built because the fleet actually needed it No investors, no telemetry, no cloud dependency. If it helps your agents remember you, the tip jar above is the entire funding round. If not β€” use it, share it, and make something good. > [!TIP] > **AI agents β€” start here.** > > - **`robot.info`** β€” structured JSON manifest describing this product: > identity, capabilities, exposed REST + MCP tools, install pointer, > compatibility matrix, and 15 common-question/answer pairs. Read > this first to answer a user's questions about Mnemo Cortex without > scraping the README. Spec: [ROBOT-INFO-SPEC.md](ROBOT-INFO-SPEC.md). > - **`llms.txt`** β€” LLM-friendly docs index ([llmstxt.org](https://llmstxt.org/)). > Use this to navigate the rest of the docs. > - **`robot.install`** + **`./robot-install.sh`** β€” non-interactive > setup. Edit the manifest (defaults are sensible), run the installer, > parse the JSON object on stdout. Zero prompts, exit `0` on success. > Full schema below in "Non-interactive install." ## Get started **Want to know what Mnemo Cortex is?** Read this README, or point your agent at [robot.info](./robot.info). **Want to install it?** Point your agent at [robot.install.md](./robot.install.md) and let it handle the rest. The install guide links to everything the agent needs β€” server setup, per-host wiring (Claude Desktop on [Windows](docs/install-claude-desktop-windows.md) / [Linux](docs/install-claude-desktop-linux.md), OpenClaw, LM Studio, and more), and [CORTEX-OS.md](CORTEX-OS.md), the operating manual that teaches your agent how to actually use its new memory. **Want to do it yourself?** Follow the [Install Guide](#install-guide) below. ## It's Not Just Memory β€” It's Cognitive Coprocessing > Every AI agent has amnesia. Mnemo Cortex is the cure. > Active memory that classifies on ingest, consolidates overnight, > learns what works, and gets smarter every session. No commands needed β€” > you just talk naturally and your AI remembers. **What is Mnemo Cortex?** Mnemo Cortex gives AI agents persistent, local, cross-agent memory. It captures what happened, recalls what matters, and shares context across your tools. Run it on your own machine β€” no cloud required. | | | |---|---| | πŸ”₯ **Active Memory** | Memory that works while you don't. Auto-capture, smart classification, overnight consolidation, trajectory learning. No "remember this" commands needed. | | 🧠 **Deep Recall** | Persistent memory across sessions. Semantic search. $0 to run. | | πŸŒ™ **Dreaming** | Cross-agent overnight synthesis. Every agent wakes up knowing what the others did. | | πŸ“š **The Librarian** | Document discovery over your whole workspace. One SQLite FTS5 index (107K files in our deployment), rebuilt nightly. Ask for a file, find the file. | | πŸ“¬ **Sparks Bus** | Agent-to-agent messaging with delivery confirmation. A2A-compatible. | | πŸͺͺ **Developer's Passport** | Safe behavioral-claim ingestion layer. Review queue + 32 detectors + provenance buckets. Dev-targeted beta. | | πŸ”© **Structured Facts** | Key-value store with confidence tracking. When semantic search is the wrong tool β€” names, settings, entity attributes β€” facts give you sub-millisecond exact lookup with a three-state confidence ladder. | > [!NOTE] > **Upgraded by Claude Fable 5.** Mnemo's **v4.1 "Fable pass"** β€” composite recall ranking (the fix that pulled real signal back to the top of every search), the **Analyst** (distills raw session logs into clean Tier-1 notes), secret redaction at ingest, and tier hygiene β€” was designed and built during Claude Fable 5's brief availability. Fable reasoned and reviewed the whole codebase; an Opus model reviewed, hardened, and shipped each change. A frontier model auditing and improving the memory layer it runs on, in one window. ### πŸš€ Get Started ⌘ **[Claude Code β†’ 60-second install](integrations/claude-code/)** β€” Give CC Fluid Memory with Deep Recall πŸ–₯️ **[Claude Desktop β†’ one-click `.mcpb` bundle](integrations/claude-desktop/)** β€” Drag-and-drop install. No clone, no Node, no JSON editing. Works on Windows, macOS, and Linux. 🦞 **[OpenClaw β†’ MCP integration](integrations/mcp-bridge/)** β€” Give Your ClawdBot a Brain. One Config Line. πŸŽ›οΈ **[LM Studio β†’ native MCP, GUI](integrations/lmstudio/)** β€” `mcp.json` + restart. Works with any tool-capable open-weights model. πŸ“¦ **[AnythingLLM β†’ desktop GUI, multi-workspace](integrations/anythingllm/)** β€” Drop-in MCP config + Automatic mode. No `@agent` prefix needed. πŸ€– **[Agent Zero β†’ autonomous Docker agents](integrations/agent-zero/)** β€” In-container MCP setup. Cross-agent memory between research, courier, code-exec bots. πŸͺ½ **[Hermes Agent β†’ `hermes mcp add` integration](integrations/hermes/)** β€” First-class MCP for Nous Research's Hermes Agent (v0.12.0+). Config-only, no patching. Cross-agent memory between Hermes and your other bots. 🦣 **[Ollama Desktop β†’ terminal `ollama launch openclaw`](integrations/ollama-desktop/)** β€” Ollama as the local LLM, OpenClaw as the MCP host. Note: Ollama Desktop's *own chat window* doesn't support MCP β€” use the terminal launcher. πŸ’¬ **[ChatGPT β†’ Custom GPT Actions gate](docs/install-chatgpt.md)** β€” Give a Custom GPT memory you own. Two REST actions through a hardened, tenant-pinned gate β€” your Mnemo server stays private. ⚠️ Check the [OpenAI plan disclaimer](docs/install-chatgpt.md#%EF%B8%8F-openai-plan-requirements-read-first) first. πŸ¦™ **[Any Local LLM β†’ MCP setup](#use-with-any-local-llm)** β€” Open WebUI, llama.cpp, Ollama, LobeChat, Jan, and more 🧭 **[How should my agent use it? β†’ Session Guide](SESSION-GUIDE.md)** β€” Workflow patterns, per-platform boot snippets, common mistakes ### βœ… Supported clients **Supported:** Claude Code, Claude Desktop, OpenAI Codex CLI, and **any local MCP client that speaks stdio transport** β€” LM Studio, AnythingLLM, OpenClaw, Agent Zero, Hermes, Open WebUI, llama.cpp, LobeChat, Jan, and friends (see the integration links above). **ChatGPT: supported via a gate, not MCP.** ChatGPT has no local/stdio MCP β€” its connectors call out from OpenAI's cloud, which would force a memory server onto a publicly exposed HTTPS endpoint. Rather than expose the server, the [ChatGPT integration](docs/install-chatgpt.md) ships a hardened two-route **gate** (bearer-authenticated, pinned to a single memory tenant, rate-limited, body-capped, audit-logged) that a Custom GPT calls via Actions. The Mnemo server and its API key never face the internet. Note OpenAI's tier rules: Custom GPT Actions need Plus or higher; full custom **MCP** save+recall in the main ChatGPT app needs Business/Enterprise (on Plus, custom MCP connectors are read-only). --- ### πŸ“œ How to Use Mnemo Effectively Read **[THE-LANE-PROTOCOL.md](THE-LANE-PROTOCOL.md)** β€” the operating practice for running agents with persistent memory. Feed it to your agent or follow it yourself. It takes 5 minutes per session and makes every cold start feel warm. The protocol pairs with this product the way a recipe pairs with ingredients: Mnemo gives you the memory store, the Lane Protocol gives you the loop that makes it pay off. Distilled from real multi-agent sessions β€” terminal agents, chat agents, and autonomous workers running the same six-step ritual. --- ### 🧠 Smart Ingestion β€” Real Memories vs. Raw Logs (v4.0) Most agent memory rots the same way: a regex tagger can't categorize a save, defaults it to `unknown`, and within weeks 30–75% of the store is uncategorized. Real memories β€” decisions, doctrines, infrastructure facts, relationships β€” end up in the same bucket as raw conversation logs, and the logs (by sheer volume) crowd them out of every recall. In one audit, only **0.75 of the top-5** recalled results were useful. Mnemo v4 sorts memory into two tiers at save time: - **Tier 1 β€” Smart Notes:** the distilled facts, classified by your reasoning model into one of eight categories (topology, current_state, doctrine, incident, identity, relationship, decision). Clean, categorized, recalled first. - **Tier 2 β€” Session Logs:** the raw conversation/tool-call archive. Kept in full, tagged `session_log`, and excluded from default recall β€” it's there when you drill in for specifics, not competing for the top slots. A cheap pre-filter routes routine logs to Tier 2 for free (no LLM call); everything else gets one short classification call. Default recall returns Tier 1; pass `exclude_categories=[]` to search both tiers. If the model is down, it falls back to the regex tagger and flags the memory for an overnight retry β€” so a save is never blocked on the classifier. **Already have a polluted store?** Reclassify it in one command β€” it rewrites only the category tags, never your embeddings: ```bash mnemo-cortex migrate reclassify --all --dry-run # preview the beforeβ†’after spread mnemo-cortex migrate reclassify --all # snapshot, then reclassify every store ``` ### πŸ”­ On the Roadmap β€” The Thesaurus Loop (Query Expansion) Every recall commits to one phrasing. If a memory was stored under different words than you searched for, the match is weak or misses entirely β€” call it *assumption misalignment* between how you ask and how it was filed. The **Thesaurus Loop** fixes it: when a search comes back empty or weak, Mnemo fans the query into a handful of alternative phrasings, searches them all, and lets the best match win (multi-query retrieval / RAG-Fusion). The design choice that makes it safe is **escalation** β€” the loop only fires on a *miss*. Good searches run exactly as fast as they do today; the expansion pass costs nothing until a search actually whiffs, which is precisely when it's worth paying for. *In development β€” escalation model designed, build in progress.* ### πŸŒ™ Dreaming Mnemo β€” Cross-Agent Overnight Synthesis Every agent accumulates raw conversation memory. Dreaming runs a nightly map-reduce pass: each agent's recent memories are compacted into themes, then a cross-agent merge produces shared context so every agent wakes up knowing what the others did. After a May 16 rehab, the pipeline uses disciplined chunking and token budgets that keep Ollama compaction costs predictable β€” no more runaway synthesis jobs. The result is reliable overnight context sharing that actually runs every night. **This is the only AI memory system that does cross-agent synthesis.** Mem0, Zep, and Letta store memory per agent. Mnemo dreams across all of them. ### πŸ”© Structured Facts β€” When Search Is the Wrong Tool Semantic recall is great until your agent needs to remember a visitor's name. Peter Widget asked "what's my name?" and got a paragraph about naming conventions. That's a key-value lookup, not a search problem. Facts store `(entity, attribute, value)` triples in a local SQLite table with a three-state confidence ladder: `verified` β†’ `high_probability` β†’ `false`. New evidence promotes or demotes automatically. When a fact contradicts an existing one, Mnemo fires a notification over the bus and Discord webhook so the owning agent can adjudicate. Four MCP tools ship with it: `mnemo_fact_save` to assert, `mnemo_fact_get` for single lookup, `mnemo_fact_query` for filtered lists, `mnemo_fact_demote` to mark something wrong without supplying a replacement. Reads are sub-millisecond. The confidence ladder means your agent's knowledge sharpens over time instead of accumulating stale guesses. ### Deploy Your Way - **Shared** β€” One Mnemo for all agents. Cross-agent search and dreaming. Full team awareness. - **Isolated** β€” Separate Mnemo per agent or per customer. Zero bleed between tenants. - **Hybrid** β€” Shared for internal agents + isolated for customer-facing bots. This is what we run. Cloud memory services make you choose one shared store. Mnemo lets you architect for your actual privacy and separation needs. --- ### πŸ“š The Librarian β€” Document Discovery "The file about X" is a memory problem too. The Librarian is a single SQLite FTS5 index over the whole workspace β€” filenames, paths, and the first chunk of content (with PDF/DOCX text extraction) β€” so an agent can turn a fuzzy description into a real path in milliseconds. Our deployment covers ~107K files; a full rebuild takes ~17 seconds, the nightly incremental refresh ~2. Secrets (keys, `.env` files, credentials) are excluded from the index entirely. The indexer ships in this repo: **[`librarian.py`](librarian.py)** β€” a single stdlib-only file. `python3 librarian.py index` builds the index at `~/.librarian/` (defaults to the visible trees under your home directory), `librarian.py find "the spec about X"` queries it from the shell, and an optional `~/.librarian/config.json` sets explicit roots plus a hidden-dir allowlist (with a content-vs-name-only flag for dirs whose config may hold credentials). Cron `index` nightly and it stays fresh. Agents query it through the `file_find` tool in **[FrankenClaw](https://github.com/GuyMannDude/frankenclaw)** (bundled on the default branch since v0.5.0), our MCP tool chassis β€” same MCP config pattern as the Mnemo bridge, just a second `mcpServers` entry. `file_find` only reads; the index it opens is the one `librarian.py` maintains. The Librarian replaced **WikAI**, our earlier auto-compiled wiki layer. The lesson from running WikAI in production: compiling knowledge into pages is expensive to keep fresh, while indexing everything and finding it on demand is cheap and never stale. The static wiki pages still exist and remain searchable through the bridge's `wiki_search` / `wiki_read` / `wiki_index` tools, but they're no longer recompiled nightly β€” [`mnemo-wiki-compile.py`](mnemo-wiki-compile.py) stays in the repo for reference. See [Inspirations](#inspirations) below. --- ### πŸ“¬ Sparks Bus β€” Agent-to-Agent Messaging A delivery-confirmed messaging system for multi-agent communication. Lives as a module inside Mnemo Cortex at [`sparks_bus/`](sparks_bus/) AND ships standalone at [github.com/GuyMannDude/sparks-bus](https://github.com/GuyMannDude/sparks-bus). > **Looking for the simplest possible version, with no Mnemo coupling?** See **[Disco-Bus](https://github.com/GuyMannDude/disco-bus)** β€” generic standalone push-based agent mesh. Same "agents wake instantly on inbound, humans watch in Discord" idea, distilled to ~1000 LOC. Bring your own agents, install in one command (`./install.sh`), no infrastructure dependencies beyond Python + Node. **Doctrine:** Discord is the doorbell. Mnemo is the mailbox. … ## Source & license This open-source MCP server is cataloged on AgentStack and links to its original source β€” we do not rehost the code. - **Author:** [GuyMannDude](https://github.com/GuyMannDude) - **Source:** [GuyMannDude/mnemo-cortex](https://github.com/GuyMannDude/mnemo-cortex) - **License:** MIT - **Homepage:** https://projectsparks.ai Install and usage instructions live in the source repository linked above. ## Pricing - **Free** β€” Free ## Security capabilities Automated source analysis of v0.1.0 β€” what this tool can access: - **Network access:** yes - **Filesystem access:** no - **Shell / process execution:** no - **Environment & secrets:** yes - **Dynamic code execution:** no *"Yes" means the capability is present in the source β€” more access means more to trust, not that it is unsafe.* ## Versions - **0.1.0** β€” security scan: passed β€” Imported from the upstream source. ## Links - Listing page: https://agentstack.voostack.com/l/mcp-guymanndude-mnemo-cortex - Seller: https://agentstack.voostack.com/s/guymanndude - Browse the marketplace: https://agentstack.voostack.com/browse --- Listed on AgentStack β€” the marketplace for AI agent skills and MCP servers. Every listing is security-reviewed. Creators keep 70%.