Long-term memory for AI agents: local-first PKM with provenance and hybrid search.
Long-term memory for your AI agents — local-first, with provenance.
Tacitus is an MCP server that turns any folder of Markdown notes into an agent-native knowledge base. It gives AI agents (Claude Code, Claude Desktop, and any MCP client) three things they actually need:
get_note discloses progressively
(outline → frontmatter → full); the wikilink graph is a queryable API.
Hybrid lexical + semantic search, with an optional neural embedder.Notes stay as plain .md files in your folder. No cloud, no lock-in.

Recorded from a real session — the memory id, changeset id, version id and audit line above are what the binary actually returned (how).
npx -y @dashiro/tacitus-mcp-server /path/to/your/vault
claude mcp add tacitus -- npx -y @dashiro/tacitus-mcp-server /path/to/your/vault
claude_desktop_config.json){
"mcpServers": {
"tacitus": {
"command": "npx",
"args": ["-y", "@dashiro/tacitus-mcp-server", "/path/to/your/vault"]
}
}
}
Prefer a single, zero-dependency binary? The Rust server ships prebuilt for macOS, Linux, and Windows on every release.
# macOS / Linux — installs `tacitus-mcp` into your Cargo bin dir
curl --proto '=https' --tlsv1.2 -LsSf \
https://github.com/ionasrobert/tacitus-mcp-server/releases/latest/download/tacitus-mcp-installer.sh | sh
# Windows (PowerShell)
irm https://github.com/ionasrobert/tacitus-mcp-server/releases/latest/download/tacitus-mcp-installer.ps1 | iex
Or grab a .tar.xz / .zip for your platform from the
latest release.
Then point any MCP client at the binary instead of npx:
claude mcp add tacitus -- tacitus-mcp /path/to/your/vault
The native binary is the flagship server (25 tools; the npm server has the
core 16 — see the table below). Both share the same on-disk formats, so a
vault works with either. Set TACITUS_SCOPE=read-only to run the native
server without write permissions.
| Group | Tools |
|---|---|
| Memory | remember, recall, forget |
| Retrieval | search, get_note, graph_query, list_notes, properties_query* |
| Write-back | propose_changes, commit_changes, revert, rename_note, delete_note, get_version* |
| Convenience | create_note, update_note, link, tag, audit_log |
| Templates | list_templates, create_from_template |
| Tasks | list_tasks, toggle_task |
| Meta | capabilities |
* Native-Rust-server first (the npm server will catch up):
properties_query — Bases-like structured queries over YAML frontmatter
(filters eq|ne|contains|exists|not_exists|gt|lt|gte|lte, sort, select,
token_budget). Templates — Markdown files in .tacitus/templates/ whose
{{var}} placeholders form a schema; substitution happens before YAML
parsing so numeric vars stay typed, {{date}}/{{time}}/{{datetime}}
auto-fill, and creation is versioned + audited like any agent write.
Tasks — every checklist line (- [ ]) as a typed entity (done, due from
due:YYYY-MM-DD or 📅, #tags), queryable and toggleable; toggling takes
the task text as a concurrency guard so a stale caller gets a CONFLICT
instead of flipping the wrong task. rename_note retargets every wikilink
that resolves to the note (alias/heading kept) in one atomic changeset —
a single revert undoes the whole rename; delete_note is versioned too.
Every tool validates input with a schema and returns structured, actionable
errors ({ code, reason, suggestion }) rather than stack traces.
In Tacitus, a plugin is an MCP client — the tool contract above is the public API, with permission scoping, versioning, and audit built in.
@dashiro/tacitus-sdk: every tool as
a typed method, {code, reason, suggestion} errors thrown as
TacitusToolError:
const tacitus = await TacitusClient.spawn({ vault: '/path/to/vault' });
const hits = await tacitus.search({ query: 'client X', token_budget: 500 });
tacitus-plugins runs
guest wasm under Wasmtime with manifest-declared permissions (tool allowlist
tacitus.call is tools/call.
The native binary embeds the runtime: tacitus-mcp plugin list|run for cron
agents and scripts.
See docs/PLUGINS.md §5TACITUS_EMBEDDER=ollama uses a local Ollama daemon
for embeddings (TACITUS_OLLAMA_EMBED_MODEL, default nomic-embed-text;
needs an Ollama with embedding support). Vectors cached in .tacitus/vectors/;
falls back to the deterministic hashing embedder when unavailable.tacitus-mcp sync init; sync status shows how much of
the relay quota a vault uses).tacitus/ internals, stable ids, note conventions)
Left: two devices, plain Markdown. Right: everything the relay stores for that
vault. Those blobs were read out of a real log.jsonl after the recording —
no note title, no path, no plaintext. The vault code is the key and never
leaves your devices; lose it and the relay's copy is undecryptable forever.
search defaults to hybrid mode (lexical + a deterministic, offline
embedder that catches morphological variants). For synonym/paraphrase matching,
opt into a neural embedder:
npm i @huggingface/transformers
TACITUS_EMBEDDER=transformers npx @dashiro/tacitus-mcp-server /path/to/vault
Vectors are cached under .tacitus/vectors/. Falls back to the deterministic
embedder if the optional dependency or model isn't available.
your-vault/
├── notes... ← your Markdown files (untouched format)
└── .tacitus/
├── memory/*.md ← agent memories (Markdown + YAML frontmatter)
├── vectors/*.json ← cached embeddings
├── history/*.json ← version snapshots (for revert)
└── audit.log ← JSONL log of every agent write
Polyglot monorepo. The reference server (shipped on npm) is TypeScript in
packages/mcp-server. A native Rust server in crates/ provides a
single-binary, zero-runtime-deps build (crates/tacitus-core engine +
crates/tacitus-mcp rmcp server) — a superset of the TS server (25 vs 16
tools). Its stable_id matches the TS engine byte-for-byte, so memory ids are
identical across both engines.
# TypeScript server
npm ci
npm test # vitest
npm run typecheck
npm run lint
npm run build # tsup → packages/mcp-server/dist
npm run eval # retrieval quality report
# Rust server (native, single binary)
cargo test
cargo clippy --all-targets -- -D warnings
cargo fmt --check
cargo run -p tacitus-mcp -- /path/to/vault # runs the MCP server on stdio
cargo build --release # → target/release/tacitus-mcp
Cross-platform release binaries are built and published to GitHub Releases by
cargo-dist (dist-workspace.toml +
.github/workflows/release.yml) on every v* tag.
MIT — see LICENSE.
Source-derived launch command. Check the maintainer’s required arguments and credentials before running:
npx -y @dashiro/tacitus-mcp-serverMerge this template into ~/Library/Application Support/Claude/claude_desktop_config.json. Keep existing servers. Add any arguments, credentials, and permissions required by the maintainer; this template has not been install-tested.
{
"mcpServers": {
"io-github-ionasrobert-tacitus-mcp-server": {
"command": "npx",
"args": [
"-y",
"@dashiro/tacitus-mcp-server"
]
}
}
}Restart Claude Desktop completely for changes to take effect. Confirm the server appears connected in the client’s tool list, then try a read-only example from its documentation.
Claude Desktop setup reference@dashiro/tacitus-mcp-servernpmio.github.ionasrobert/tacitus-mcp-server works with any MCP-compatible client. Copy the config snippet from the Configuration section above and add it to the file shown for your client, then restart the application.
~/Library/Application Support/Claude/claude_desktop_config.jsonRestart Claude Desktop completely for changes to take effect.~/.cursor/mcp.jsonRestart Cursor for changes to take effect..vscode/mcp.jsonReload VS Code window for changes to take effect.~/.codeium/windsurf/mcp_config.jsonRestart Windsurf for changes to take effect..mcp.jsonSave at the project root, then start Claude Code in that project and review the MCP server approval prompt. Keep real credentials out of shared files.