MCP.Market

Tacitus MCP server

by ionasrobert·io.github.ionasrobert/tacitus-mcp-server·v0.1.3
No reviews yet

Long-term memory for AI agents: local-first PKM with provenance and hybrid search.

A92/100grade A
Write a reviewInstall
What users say
No reviews yet
Be the first
Safety scan
A92/100

full report

Adoption
Growing

0 stars226 downloads/wk

Reviews

Write one

Nobody has reviewed Tacitus yet.

If you have run it, two minutes of your experience saves the next person an afternoon.

Tacitus tools (16, 3 write)

write = sends, deletes, buys or posts

Read from the package source without running it. The installed server may list more.

  • audit_log

    Read recent agent actions (commits/reverts), most recent first.

  • capabilities

    List available tools and the current permission scope.

  • commit_changes

    Atomically apply a previously proposed change_id (all-or-nothing). Returns a version_id usable with revert.

  • create_notewrite action

    Create a note (auto-committed). Fails if it already exists. Returns a version_id.

  • forgetwrite action

    Delete a memory by id. Returns whether a memory was removed.

  • get_note

    Fetch a note progressively: outline (headings) | frontmatter_only | full, with an optional max_tokens ceiling.

  • graph_query

    Traverse the wikilink graph: links (outgoing) | backlinks | neighbors (both directions, to depth).

  • link

    Append a [[to]] wikilink to the `from` note (idempotent). Returns a version_id.

  • list_notes

    List all note ids with their titles and paths.

  • propose_changes

    Dry-run a changeset (create/update/delete notes). Returns a change_id and a before/after diff. Nothing is written until commit_changes.

  • recall

    Recall memories relevant to a query, ranked, within an optional token_budget. Surfaces conflicting memories instead of silently choosing.

  • remember

    Store a typed memory (user|feedback|project|reference) with mandatory provenance. Returns a stable memory_id; idempotent for identical content+source.

  • revert

    Undo a committed change by version_id, restoring the prior note state.

  • search

    Search vault notes by relevance; returns ranked snippets within an optional token_budget (never whole notes). mode: hybrid (default) | lexical | semantic. Expand with get_note.

  • tag

    Add a tag to a note (deduplicated). Returns a version_id.

  • update_notewrite action

    Update a note body and/or frontmatter (auto-committed). Returns a version_id.

Public scan report

scanner v0.1.10 · 2026-11-03 · same rubric, same numbers if you re-run it

no findings
  • Code scan2 source files scanned25/25
  • –Live reliabilityno gateway calls yet and no remote to proben/a
  • –Tool poisoningtools not inspected (local package is not executed); not countedn/a
  • Auth qualitylocal package, no credentials required12/15
  • Maintenancelast push 4 days ago15/15
  • Maintainer identityregistry namespace matches repository owner; GitHub account older than a year8/10
Overall 92/100. Components that don't apply are left out of the denominator. Any critical finding is an F.RubricAppeal a findingJSON

How Tacitus compares

Against every graded Web search MCP server, from our own scans and uptime checks. Updated after each scan run.

Safety score
92/100
higher than 90% of 3,470 Web search servers
Tools
16
the typical Web search server has 5
Tools that change things
19%
typical: 0%. Fewer write tools means less an agent can break
Downloads a week
226
more than 72% of 827 Web search servers with a package

What its tools let an agent do: 13 read · 1 create · 2 change

Release history: 2 versions in the registry since 2026-08-14, latest 2026-10-31, typically one every 78 days.

Install Tacitus in Claude Code, Cursor or VS Code

Runs npx -y @dashiro/tacitus-mcp-server on your machine. Read the scan report first; the gateway never runs local packages.

claude mcp add tacitus-mcp-server -- npx -y @dashiro/tacitus-mcp-server
Add to Cursor

What the publisher says

From the Tacitus repository's README, as published. We do not edit it. Read it on GitHub

Tacitus

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:

carries its source and is returned within a token budget; contradictions are surfaced, not silently resolved.

  1. Memory with provenance — typed, queryable long-term memory. Every fact

(never whole notes) under a token budget; get_note discloses progressively (outline → frontmatter → full); the wikilink graph is a queryable API. Hybrid lexical + semantic search, with an optional neural embedder.

  1. Retrieval that fits the context window — search returns ranked snippets

atomically, and revert by version. Read-only scope forbids mutations; every write is audited.

  1. Safe write-back — propose a changeset, preview the diff, commit

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).

Quick start

npx -y @dashiro/tacitus-mcp-server /path/to/your/vault

Claude Code

claude mcp add tacitus -- npx -y @dashiro/tacitus-mcp-server /path/to/your/vault

Claude Desktop (claudedesktopconfig.json)

{
  "mcpServers": {
    "tacitus": {
      "command": "npx",
      "args": ["-y", "@dashiro/tacitus-mcp-server", "/path/to/your/vault"]
    }
  }
}

Native binary (no Node)

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.

Tools

Shortened. The full README is on GitHub.

Nothing above is checked by us. What we check is on the safety report.

Tacitus: common questions

Is Tacitus MCP server safe?
Yes, by our scan: it is graded A (92/100). Read the Tacitus safety report
How do I install Tacitus?
It runs on your machine. Copy the Claude Code, Cursor, VS Code or Claude Desktop config from the install section.
Does Tacitus need an API key?
Not as far as the registry entry and our scan can tell: no credentials are declared or required.
Is Tacitus maintained?
The last commit was 4 days ago (2026-10-31). The latest release is v0.1.3.
What can I use instead of Tacitus?
Servers from other publishers that do the same job: M3 Memory MCP server, Reverie MCP server and Pubmed MCP server.
How does Tacitus compare with other Web search MCP servers?
safety score 92/100 (higher than 90% of 3,470 Web search servers); tools 16 (the typical Web search server has 5); tools that change things 19% (typical: 0%. Fewer write tools means less an agent can break). Full comparison.

Alternatives to Tacitus

Same job from other publishers: the closest match first, then the best rated.

  • M3 Memory
    Local-first, cloud-friendly memory — 100+ tools, 99.2% LongMemEval-S retrieval@8, hybrid search.
    A
  • Reverie
    Graph memory that dreams: Neo4j knowledge-graph memory for AI agents with hybrid search
    B
  • Pubmed
    Search PubMed/Europe PMC, fetch articles and full text (PMC/EPMC/Unpaywall), citations, MeSH terms.
    A
  • Jshookmcp
    Search-first MCP server for JavaScript reverse engineering: 735 tools across 36 domains.
    C
  • DocsMint
    Cloud or self-hosted knowledge for AI agents: hybrid search, reranking, GraphRAG, scoped MCP tools.
    A