NeuroStack MCP server

A local retrieval layer and optimizer for the Markdown knowledge base you already have.

NeuroStack indexes a folder of .md files (Obsidian, Logseq, Notion exports, plain Markdown) into one SQLite database with full-text search, embeddings and a wiki-link graph. Any MCP client can then search it, walk the graph and save memories. A query returns ranked evidence with note paths, and your AI does the reasoning, so no model runs while you wait. Between queries NeuroStack keeps the knowledge base accurate. It flags notes that have gone stale, harvests decisions and root causes from your AI sessions, merges recurring memories into learnings, and queues the proven ones to become notes. Indexing never changes your files, and optional MCP write tools let a client edit notes through your git history.

Works with Claude, Cursor, Windsurf, Gemini CLI, VS Code, Codex and any other client that supports MCP.

60-second start

npm install -g neurostack          # install
neurostack init                    # pick your notes folder, press Enter for the rest
/save                              # in omp or Claude Code, keep what this session learned
neurostack doctor                  # check that everything works

The steps below explain each part and how to connect other AI clients.

---

Your notes, in your control

By default, NeuroStack is a read-only indexing layer:

• Indexing, search, summaries, and graph analysis never modify your Markdown files
• All index data lives in NeuroStack's own separate database
• To remove it completely, run neurostack uninstall. Your notes stay untouched.
• Nothing ever leaves your machine, unless you configure a third-party LLM provider for summaries and embeddings

If your vault is a git repo, four opt-in MCP write tools let an AI client author and edit notes for you: vault_write_file, vault_delete_file, plus vault_read_file / vault_list_files. Every write commits and pushes to your git remote with a descriptive message, so every change is visible in git log, revertable with git revert, and serialised under a per-vault lock. Writes hard-reject invalid frontmatter, paths outside the vault, and hidden directories (.git, .obsidian, …). Because the tools are exposed to any client talking to neurostack serve, gate them at the transport (auth, tunnel, LAN only) if you put the MCP endpoint on the public internet.

---

Who this is for

You do not need to be a developer. If you take notes in Markdown, or can export your notes as Markdown from Obsidian, Notion, Bear, or Roam, NeuroStack works for you.

If you are...	NeuroStack helps you...
A researcher	Ask your AI "what do my notes say about X?" across hundreds of papers. Get warned when a note references a retracted finding or superseded paper before your AI cites it confidently.
A fiction writer	Your AI knows your world-building bible, character histories, and chapter decisions. It remembers that you agreed in session 4 that Elena's backstory changes in act 2.
A student	Ask your AI to explain connections across all your course notes. When a syllabus topic changes, stale revision notes are flagged automatically.
A professional	Your AI remembers client context, project decisions, and meeting notes session-to-session. No more re-pasting the same background every time.
A developer or DevOps engineer	Notes that reference deprecated APIs or reversed architecture decisions get flagged before your AI cites them as current.

---

Get started in three steps

You will need Node.js installed (most computers already have it). The npm package handles the Python setup for you.

Step 1. Install

npm install -g neurostack

Step 2. Set up (takes about two minutes)

neurostack init

The setup wizard asks which vault folder to index, which mode to run (Lite or Full), and which profession pack to apply. Then it finishes the install for you. It adds the NeuroStack hook to omp or Claude Code if you have them, so /save keeps what a session learned. It sets up the AI that turns saved sessions into memories, using the same model as your summaries. If no model answers yet, it leaves that off and tells you so. It also installs one background timer that keeps the index tidy, and a failed background job shows up as a desktop notification. Press Enter at each question to take the default. At the end it runs neurostack doctor and shows what to try first.

Step 3. Connect to your AI

For Claude Desktop:

neurostack setup-desktop

For Claude Code:

claude mcp add neurostack -- neurostack serve

For Cursor, Windsurf, Gemini CLI, or VS Code:

neurostack setup-client cursor      # or: windsurf, gemini, vscode

Done. Open a new conversation and ask your AI about something from your notes.

Everything runs on your machine. Choose a tier during neurostack init:

• Lite (~130 MB) gives you keyword search, link-based connections between notes, stale detection and the MCP server. No GPU or Ollama required.
• Full (~560 MB) adds semantic search by meaning, AI-generated summaries, connections between notes, and topic clustering via local Ollama. GPU or 6+ core CPU recommended.

Non-interactive setup takes every default:

neurostack init --mode lite ~/my-notes --yes    # lite mode
neurostack init --mode full ~/my-notes --yes    # full mode

--no-hooks, --no-schedule and --checkpoint none turn off the hook, the timer and the checkpoint AI.

# PyPI
pipx install neurostack
pip install neurostack        # inside a venv
uv tool install neurostack

# One-line script
curl -fsSL https://raw.githubusercontent.com/raphasouthall/neurostack/main/install.sh | bash

# Lite mode (no ML deps)
curl -fsSL https://raw.githubusercontent.com/raphasouthall/neurostack/main/install.sh | NEUROSTACK_MODE=lite bash

On Ubuntu 23.04+, Debian 12+, and Fedora 38+, bare pip install outside a virtual environment is blocked by the operating system. Use npm, pipx, or uv tool install instead.

To uninstall: neurostack uninstall

---

What it does

• Hybrid search (FTS5 keyword + semantic) with tiered depth, so a client can fetch triples, summaries or full notes by token budget.
• Ranked evidence with note paths and excerpts over the CLI, MCP, or an OpenAI-compatible API, for your AI to cite and reason over.
• Stale detection. A note that keeps surfacing in contexts where it no longer fits is flagged and demoted in later results.
• Session harvest. A timer scans Claude Code, Codex, Gemini and omp transcripts and saves decisions, bugs, conventions and learnings as memories with TTLs.
• Synthesis and promotion. Recurring memories become learnings; a promotion queue lists which ones are ready to become notes.
• Wiki-link graph with PageRank, community detection, gap and bridge analysis.
• Read-only by default. Opt-in write tools commit and push every change to your git remote.

Editable sources live in the .drawio files next to the images.

---

Design

Philosophy

1. Your files stay yours. The index lives in its own database, so deleting it leaves your notes exactly as they were.
2. NeuroStack returns evidence and your AI reasons over it. No tool answers questions, and tests/test_no_answering.py keeps it that way.
3. Removing a memory moves it to an archive that search cannot reach, and you can restore it later.
4. NeuroStack redacts secrets from text a model wrote, such as harvested summaries. It stores what you or your agent saved on purpose word for word.
5. Ranking changes get measured first. neurostack eval switches off usage learning while it runs, so one pass cannot skew the next.

Mental model

NeuroStack holds two layers and one loop.

The vault is your Markdown. NeuroStack reads it, splits it into chunks, embeds them, summarises each note, extracts facts as triples and builds the link graph. Note status and tags live in NeuroStack's own note_metadata table, so it never needs to edit your frontmatter.

The memory layer holds what agents learn, typed as observations, decisions, conventions, learnings, bugs or context. These rows belong to NeuroStack and carry an optional workspace and an optional expiry.

The optimizer loop runs on a timer and connects the two. A memory usually starts in an AI session, becomes a learning when it keeps recurring, and becomes a note once no existing note covers it.

Architecture

flowchart LR
    V[/"Markdown vault<br/>never modified"/] --> IX["index<br/>chunk, embed, summarise,<br/>triples, link graph"]
    IX --> DB[("neurostack.db<br/>SQLite + FTS5")]
    S[/"AI session transcripts"/] --> H["harvest"]
    A["MCP client or CLI"] -->|vault_remember| DB
    H --> DB
    DB --> R["search, tiered, context,<br/>graph, brief"]
    R --> A
    DB --> J["timer jobs<br/>synthesize, promotion,<br/>decay, communities"]
    J --> DB
    J -->|opt-in, git commit| V
    DB --> UI["neurostack ui<br/>read-only dashboard"]

One OS timer runs neurostack run-due every minute, and that command runs each due job in turn. neurostack serve exposes 33 MCP tools over stdio or HTTP, and neurostack api serves the same retrieval as an OpenAI-compatible API.

Data model

Area	Tables	Purpose
Vault index	notes, chunks + chunks_fts, summaries, triples + triples_fts, graph_edges, note_metadata	Searchable copy of your notes. Deleting a note cascades to its chunks, summaries and triples.
Topics	communities, community_members, folder_summaries	Topic clusters and folder summaries for broad questions.
Memories	memories + memories_fts, memories_archive, memory_sessions	Live memories and their embeddings. The archive has no full-text index and no embedding, so search cannot reach it.
Quality signals	prediction_errors, trigger_log, memory_coverage, note_usage	Stale notes, memories that drifted from the notes they cite, ignored reminders, cached coverage verdicts, and usage for hotness.
Feedback and jobs	search_log, search_feedback, job_runs, job_queue	Implicit search feedback (opt-in), job history, and queued checkpoint work.

A memory row stores its content, tags, type, workspace, source agent, embedding and expiry. It also records embed_pending when the embedder was down, so neurostack backfill can embed it later, plus revision_count and merged_from for edits and merges. A second database, sessions.db, indexes raw session transcripts for neurostack sessions search.

Write mechanics

1. Deliberate saves. vault_remember and neurostack memories add store the text as written, embed it once, then report any near-duplicate above 0.85 similarity without refusing the save. If embedding fails, the row still saves and gets flagged for backfill.
2. Harvest. The index LLM reads each session message and keeps decisions, root causes, rules and facts. A second pass reads the session as a whole and saves up to 3 final conclusions tagged session-verdict. The judge model then picks each memory's type. Harvest redacts secrets, skips anything above 0.88 similarity to a live memory, and records how many messages of each session it has read.
3. Synthesis. Observations at least 7 days old that cluster at 0.75 similarity or higher (one anchor plus 3 or more others) become one learning. The originals stay, tagged superseded_by:<id>.
4. Removal. Forget, the losing side of a merge, expiry and prune all go through one helper that copies the row into memories_archive with a reason and a timestamp, then deletes it from the live table. Search checks expiry before it runs, so an expired memory never shows up.
5. Vault writes. Only the opt-in tools (vault_write_file, vault_delete_file) and the promotion job touch the vault, and each one commits and pushes to git.

Read mechanics

1. Hybrid search. NeuroStack runs keyword search and embedding search separately, takes the top 50 chunks from each, and merges the two lists by reciprocal rank (k = 60). A chunk found by only one search still competes. It then adds a convergence bonus, a 1.4× boost for the caller's context, usage hotness and co-occurrence, and demotes notes flagged as stale.
2. Diversity. Results keep one chunk per note, and similar notes suppress each other, so one long note cannot fill the result list.
3. Tiered depth. A client asks for triples (about 15 tokens), summaries (about 75) or full notes (about 300), or lets auto escalate only when the cheaper tier misses.
4. Context assembly. vault_context splits its token budget into 40% memories, 20% triples, 30% summaries and 10% sessions, so no single source crowds out the rest.
5. Hard limits. Every search returns at most top_k results. No retrieval path calls an LLM unless you pass rerank=True (see below).

Optional reranking. vault_search(rerank=True) asks the judge model to score each whole-note result against your query and returns the judge's order. On 76 real search clicks it raised MRR from 0.503 to 0.652, top-1 hits from 34% to 49% and top-3 hits from 57% to 75% (paired permutation p=0.004). Each search costs about $0.0004 and adds about 0.4 s. It works with depth="full" or reference_only=True and raises an error at the other depths. If the judge fails, you get the normal order back.

Usage makes a note easier to find again. It never marks a note as true.

Known limits

• Harvested memories do not record their workspace yet, so a workspace-scoped search misses them.
• A harvested memory does not link back to the session and message it came from.
• A forgotten memory can come back if a later session says the same thing, because the duplicate check reads only live memories.
• Editing a memory replaces its text, and only revision_count shows that it changed.

External services

Lite mode needs none of these. Everything below is opt-in and set in ~/.config/neurostack/config.toml, and each one has a matching NEUROSTACK_* env var. When a service is missing, the feature that needs it switches off and the rest keeps working.

Service	Config keys	Default	What uses it	Without it
Embedder, any OpenAI-compatible /v1/embeddings	embed_url, embed_model, embed_api_key, embed_timeout_s	Local Ollama, nomic-embed-text	Semantic search, memory similarity, duplicate checks, synthesis clusters	Keyword search only. New memories get flagged for neurostack backfill.
Index LLM, Ollama or any OpenAI-compatible chat API	index_llm_url, index_llm_model, index_llm_api_key, or index_llm_command for a CLI such as claude -p	Local Ollama, phi3.5	Note summaries, triples, topic labels, harvest extraction, synthesis	No summaries or triples. Harvest cannot extract memories.
Judge model, OpenRouter decisions API	judge_url, judge_model, judge_api_key	~typesafe/jev-latest on OpenRouter, fails without a key	Harvest memory types, rerank=True, the promotion queue's "uncovered" check	Harvest keeps the index LLM's type, rerank returns the normal order, and uncovered memories are counted as pending.
Agent, Pi (@earendil-works/pi-coding-agent) on OpenRouter or a compatible proxy	agent_provider, agent_model, agent_api_key, agent_base_url	anthropic/claude-sonnet-5, needs Node 22.19+	neurostack agent promotion, which writes notes from the promotion queue, and neurostack agent vault-save, which writes one session transcript into the notes	The promotion job shows as blocked. The queue still builds.
Checkpoint command, any shell command that answers a prompt on stdin	checkpoint_command	Off	Server workers that turn queued /save checkpoints and harvest jobs into memories	Queued checkpoints wait. Local harvest still runs.
Notify command, any shell command	notify_command	Off	Receives a failed job run as JSON, for example ntfy publish mytopic	Failures show only in neurostack jobs and the dashboard.
Git remote on your vault	Your vault's git config	None	vault_write_file, vault_delete_file and promotion commit and push each change	A write rolls back when the push fails. Reading never needs git.

Only the embedder and the index LLM see your note text. The judge sees queries, memories and the notes being ranked or checked. Point every URL at localhost to keep all data on your machine.

---

What makes it different

NeuroStack is not a replacement for Obsidian, Notion, or any note-taking app. It sits on top of what you already use and adds what they don't have.

Capability	Note apps	Basic RAG	NeuroStack
Stores your notes	Yes	No	No (read-only by default; opt-in git-backed write tools)
AI can search your notes	Some	Yes	Yes
Detects stale/outdated notes	No	No	Yes
AI memories persist across sessions	No	No	Yes
Works with any MCP-compatible AI	No	Varies	Yes
Tiered retrieval (saves 80-95% tokens)	No	No	Yes
Profession-specific workflows	No	No	Yes
Open source, self-hostable	Varies	Varies	Yes (Apache 2.0)

Stale detection is the part other tools lack. When a note keeps appearing in contexts where it no longer fits, such as a deprecated API or a superseded paper, NeuroStack flags it and demotes it in later results.

---

Profession packs

When you run neurostack init, you choose a profession pack. Each one configures NeuroStack with templates, folder structures, and AI guidance suited to how your profession actually uses notes.

Pack	Built for
researcher	Literature review, citation tracking, evolving arguments, stale paper detection
writer	Character sheets, world-building, chapter outlines, continuity tracking
student	Course notes, spaced repetition, exam prep, syllabus change detection
developer	Code decisions, architecture notes, runbooks, deprecated API detection
devops	Infrastructure runbooks, incident notes, change logs
data-scientist	Experiment tracking, model notes, dataset documentation

Apply a pack to an existing vault without losing any notes:

neurostack scaffold researcher ~/my-notes    # or: writer, student, developer, devops, data-scientist

You can also import an existing Markdown directory:

neurostack onboard ~/my-notes

---

How retrieval works

Most memory tools give your AI a wall of text and let it figure out what's relevant. NeuroStack is tiered. It starts with the cheapest retrieval that answers the question and escalates only when it needs to.

Level	Tokens	What your AI gets
Quick facts	~15	Structured facts extracted from your notes: experiment-3 used learning-rate 0.001
Summaries	~75	AI-generated overview of a note
Full content	~300	Actual Markdown content
Auto (default)	Varies	Starts at quick facts, escalates only if the answer isn't there

Simple factual questions resolve at ~15 tokens. Deep dives get full context. Your AI spends its attention budget where it matters.

---

Your AI remembers decisions

Across sessions, your AI can save and retrieve typed memories: observations, decisions, conventions, learnings, bugs. When you start a new session, those memories are surfaced automatically.

"We decided to keep authentication stateless." "The thesis framing shifted from consolidation to complementary learning systems." "Elena's surname changed from Vasquez to Reyes in the chapter 7 revision."

These aren't just notes. They're things your AI remembers you decided together. They survive /clear. They survive closing the terminal. They survive switching machines.

neurostack memories add "revised thesis framing to CLS, not just consolidation" --type decision --tags "thesis,neuroscience"
neurostack memories search "thesis direction"

---

Learns from your AI sessions

NeuroStack scans your past AI conversations on a timer, extracts the decisions, observations and learnings, and saves them as memories. You do not have to write them down yourself.

neurostack harvest --sessions 5          # extract insights from last 5 sessions
neurostack schedule install              # one timer runs due jobs every minute

Supports Claude Code, VS Code, Codex CLI, Aider, and Gemini CLI session formats.

---

Keeps itself current

Your vault changes. NeuroStack watches it.

neurostack watch     # auto-index on vault changes

The index updates as you write and stale detection runs continuously, so you do not maintain it by hand.

---

Scheduling

NeuroStack runs its background jobs from one OS timer. The timer calls neurostack run-due once a minute, and run-due works out which jobs are due and runs them one at a time. neurostack init installs the timer for you. Installing it records each daily job as just run, so the jobs start at their next scheduled time instead of all at once. The commands below manage it by hand.

neurostack schedule install   # install or overwrite the timer
neurostack schedule status    # show whether the timer is active
neurostack schedule remove    # remove the timer

schedule install uses the scheduler your OS already ships:

OS	Backend	What it writes
Linux	systemd user timer	~/.config/systemd/user/neurostack.timer and neurostack.service
macOS	launchd agent	~/Library/LaunchAgents/io.neurostack.run-due.plist
Windows	Task Scheduler	a task named NeuroStackRunDue

The timer calls the neurostack executable it finds on your PATH at install time, so run schedule install again after you move or reinstall NeuroStack. All three subcommands accept --json.

A systemd user timer runs only while you have a login session. On a headless Linux server, or when you run NeuroStack as root, enable lingering for that user so the timer keeps running after you log out:

sudo loginctl enable-linger <user>

neurostack doctor shows when each job last ran.