Datacron

Datacron

Local MCP server to query and maintain a Markdown vault from Claude, Codex, Gemini, or another stdio MCP client, without sending the whole vault into the context.

[License: Apache 2.0] [Python: 3.11+] [MCP: local stdio] [CI]

English | Français

What can you do with Datacron?

Recover project context, prepare for a conversation, and keep track of commitments. Datacron gives your assistant durable memory in readable, editable Markdown. Your notes remain usable independently of the client you choose.

Offline library: build a browsable Markdown library for Obsidian or another local reader, with a home page and topic indexes. Review proposed rewrites, splits and archives before changing the source vault. See the offline library guide and the 2026.0913.02 release notes.

The assistant orchestrates these requests using the available tools and granted permissions. A shared protocol guides reading, people updates, and write verification. Ambiguous identities require clarification; storing a deadline does not schedule a reminder. Explore daily follow-up.

Start here: install · first session · user guide · MCP reference · privacy.

Installation

Windows: one double-click installer

The easiest way on Windows: download Datacron-Setup.exe from the latest Release, double-click it, and pick your vault. No Python, no terminal, no administrator rights; Datacron registers itself with your AI clients automatically. Full guide: Windows installation.

Python: from PyPI

python -m pip install datacrondatacron setup

From source

From a clone of the repository:

python -m pip install -e ".[dev]"

Or, to install only the application:

python -m pip install -e .

Runtime prerequisites:

• Python 3.11+
• ripgrep available on the PATH for search_regex; without it that tool falls back to a slower scan of indexed chunk bodies
• a folder of Markdown notes
• a supported stdio MCP client, such as Claude Desktop, Codex CLI, or Gemini CLI

First session

1. Choose your notes folder with the installer or datacron setup.
2. Reconnect Datacron in your MCP client to load the tools and instructions. The Claude Desktop chat does not present the server instructions: paste the session start line printed by setup into your Claude preferences (see setup).
3. Ask: "Find the notes for my project and summarize its status with sources."

For memory sessions, session_context returns bounded context and the shared protocol. prepare_follow_up prepares sourced updates; existing writers apply them according to permissions. get_follow_up retrieves the latest structured revisions. Existing prose notes remain readable and are not automatically converted.

The server operates locally. Your client may send returned excerpts to its model provider; see privacy and security.

For cached session contracts, archive ranking, resumable write tracking and conversation evaluation, see daily workflow improvements.

Quick start

The easy path - one command detects your AI clients, initializes the vault, indexes it, and registers Datacron everywhere:

datacron setup            # interactive; add --yes for all defaults

See the installation guide for options (--client, --scope, writing, durability). Or step by step:

datacron init /path/to/vaultdatacron index --vault /path/to/vaultdatacron status --vault /path/to/vaultdatacron mcp install --client claude-desktop --vault /path/to/vault

The mcp install subcommand above is dedicated to Claude Desktop. For Codex CLI, Gemini CLI, Antigravity, LM Studio, Cursor, and the other clients, use multi-client setup with datacron setup --client <identifier> or auto-detection with --client all.

Add to LM Studio

LM Studio 0.3.17+ has one user configuration and no project scope. The preferred command is:

datacron setup --yes --vault "VAULT_PATH" --client lmstudio --scope user

For a Python installation where datacron-mcp is on PATH, the equivalent read-only configuration can also be imported with this official deeplink:

Add to LM Studio

The link imports this example. Open LM Studio's MCP editor and replace both <YOUR_VAULT> placeholders before starting the server:

{  "mcpServers": {    "datacron": {      "command": "datacron-mcp",      "args": [],      "env": {        "DATACRON_VAULT_ROOT": "<YOUR_VAULT>",        "DATACRON_READ_PATHS": "<YOUR_VAULT>",        "DATACRON_DURABILITY": "best-effort"      }    }  }}

The example does not enable write tools. CLI setup is safer for packaged installations because it writes the actual executable path automatically.

Restart the configured client or clients after installation.

To run the server manually:

datacron mcp serve --vault /path/to/vault

The direct script entry used by the installer is also available:

datacron-mcp

datacron-mcp reads the vault from DATACRON_VAULT_ROOT.

Configuration

datacron init creates .datacron/VAULT.yaml. That file can carry vault-local configuration, notably query expansion:

query_expansion:  supervision: [monitoring]  sauvegarde: [backup]  restauration: [restore]  chiffrement: [encryption]  sécurité: [security]  validité: [validity]  certificat: [certificate]

Useful environment variables:

Path lists use the OS separator (: on Unix, ; on Windows).

Writing

Writes are deliberately OFF by default. Without DATACRON_WRITE_PATHS, write tools return a clear error and create no file.

To enable writing to a specific subfolder:

$env:DATACRON_VAULT_ROOT = "C:\Notes"$env:DATACRON_READ_PATHS = "C:\Notes"$env:DATACRON_WRITE_PATHS = "C:\Notes\_memory"datacron mcp serve --vault C:\Notes

datacron setup can also apply the allowlist machine-wide (user environment variable, opt-in) so every MCP client inherits it; default: _memory, _drafts, _journal. See the setup guide.

Available write tools:

• create_note_ai: creates a typed Markdown note, without overwrite.
• append_journal: adds an entry under a heading of an existing note.
• set_frontmatter: updates lifecycle fields, the rejected options list, and the monotone last_id counter without modifying the Markdown body.
• patch_note_preamble: replaces or removes the Markdown preamble before the first recognized Markdown heading (ATX or Setext), with mandatory CAS control.
• patch_note_section: replaces the content under an existing heading with CAS control.
• delete_note_section: explicitly deletes an H2-H6 section (ATX or Setext) and its subtree.
• rename_note_section: renames only the title of an H2-H6 section (ATX or Setext).
• move_note_section: previews or commits an exact H2-H6 subtree move within a note, with mandatory CAS.
• revert_note: restores the exact bytes of a version kept in history.
• apply_organization_manifest: validates and then applies a local content-addressed bundle after confirmation bound to the exact admitted organization pre-state.

Guarantees:

• strict note confinement within DATACRON_WRITE_PATHS; organization-batch note sources and targets must also stay inside the unchanged live organization.scope and pass the live note-admission policy, including exclusions
• two internal exact-CAS targets for an organization batch: .datacron/VAULT.yaml, only to change the top-level organization mapping without changing organization.scope, and .datacron/ulids.json, only when Datacron derives the key migration required by a move_replace_exact
• atomic overwrite via temporary file + os.replace
• content-addressed history before modifying an existing note
• synchronous reconcile() after a normal write; immediate searchability is guaranteed only when reconciliation succeeds
• local audit log
• for an organization manifest: crash-consistent recovery and atomic replacement of each file; simultaneous visibility across several paths is not guaranteed

Concurrent multi-machine mode is not supported for writes: keep a single-writer rule on the vault.

For apply_organization_manifest, also stop every other Datacron client and server during the maintenance window. Before applying, keep a verified byte-exact backup outside the vault of the affected notes and the complete .datacron directory until every post-commit check is green. Call mode="validate" first, review the bounded hashes it returns, then reuse the exact confirmation_token with mode="apply". The token binds the manifest and payloads, all admitted Markdown notes inside organization.scope, the exact vault configuration and identity sidecars, and the projected report. It deliberately does not bind unrelated note bytes outside organization.scope. A change to any authenticated component invalidates the confirmation before mutation. history_mode=full is required at validation time. If Datacron derives identity-sidecar case-collision cleanup, also review identity_sidecar_case_canonicalization_count and its content-free SHA-256 before applying; both proofs are token-bound and retained in the durable receipt. An existing replace_exact or move_replace_exact source must carry its id in frontmatter; an identity available only from the sidecar is unsupported by this v1 schema. If the batch is already durably committed but index reconciliation or the planner oracle fails, the response says so explicitly (committed_index_incomplete or committed_report_mismatch) and the same call can be retried with the same token. An organization-batch blocker is reported by datacron ops inspect with a pending_batch_ reason and both single-note repair actions unavailable; use the full offline rollback procedure in the operational-health guide rather than repairing or quarantining one member.

Available capabilities

Datacron indexes a folder of Markdown notes, exposes a local MCP server, then returns the relevant notes or chunks to the client instead of a full dump. The vault stays an ordinary Markdown folder: Datacron only adds a .datacron/ sidecar for the index, logs, internal ULIDs, history, and the operation journal. The one exception is datacron setup at project scope, which is part of the default: it also writes each detected client's project config into the vault root, such as .mcp.json, .cursor/mcp.json, .gemini/settings.json, .agents/mcp_config.json, .codex/config.toml or .vscode/mcp.json. Those files carry machine-local absolute paths, so a synced vault carries them to every machine. Pass --scope user to keep the vault free of them.

MCP Tools

The tables below list the standard profile (23 tools). For assistants that choose better from a short list, DATACRON_TOOL_PROFILE=essential registers 5 tools instead: remember, recall, what_changed, get_note and get_health. See the essential tool profile.

Reading

Writing

Operational

Advisory (experimental)

MCP resources:

• datacron://vault/map
• datacron://vault/info
• datacron://policy/active

Search

search_text combines several signals:

• FTS5/BM25 for the base lexical score
• a heavier weight on the note title and heading trail than on the chunk body, so a note about a subject outranks a note that merely mentions it
• FR↔EN query expansion configured in VAULT.yaml
• conservative temporal re-rank:
  • a note referenced in another note's supersedes is strongly demoted
  • confidence: low and confidence: needs_verification apply a light penalty
  • include_superseded=true brings historical notes back up
• optional scope: folder, tags, and frontmatter narrow the searched notes with the same semantics as list_notes; the response echoes the filters actually applied
• optional grouping: group_by_note=true keeps the best chunk of each note and reports how many of its chunks matched, which cuts the returned tokens by about 40 percent on the eval corpus

search_regex stays literal: it applies neither query expansion nor temporal re-rank.

recall@5       0.89recall@10      0.95recall@20      0.95MRR            0.73nDCG@10        0.79latency p50    57 mslatency p95    276 mspayload tokens 90567

Privacy and security

• Datacron does no telemetry.
• Datacron calls no cloud LLM.
• The MCP client, for example Claude, Codex, or Gemini, may send the chunks that Datacron returns to its provider. Datacron does not send it the full vault.
• Content returned to clients is wrapped in <vault_content>...</vault_content>.
• Results are bounded by count and by token budget.
• Filesystem access is confined by DATACRON_READ_PATHS and DATACRON_WRITE_PATHS.
• MCP operations are audited in the local logs.

CLI commands

datacron companion                  # launch the local graphical control panel (127.0.0.1)datacron setup                      # guided path: init + index + client configdatacron setup --yes                # all defaults, no promptsdatacron setup --client all --scope both --vault /path/to/vaultdatacron setup --vault /path/to/docs --memory-vault /path/to/memory  # Dual-Vault setupdatacron migrate-memory --vault /path/to/docs --memory-vault /path/to/memorydatacron graph-health --vault /path/to/vault --jsondatacron compact-memory --vault /path/to/vault --older-than-days 30 --dry-rundatacron compact-memory --vault /path/to/vault --older-than-days 30 --applydatacron config get --vault /path/to/vault --jsondatacron config set --vault /path/to/vault --expected-hash sha256:... --values-json '{"excluded_folders":[".git"]}'datacron setup --protocol           # also install client memory rulesdatacron protocol install --client alldatacron protocol status --client all --scope userdatacron init /path/to/vaultdatacron status --vault /path/to/vaultdatacron index --vault /path/to/vaultdatacron reindex --vault /path/to/vaultdatacron scrub-init --vault /path/to/vaultdatacron scrub --vault /path/to/vaultdatacron reorganize --vault /path/to/vault --dry-run          # measure organization, read-onlydatacron reorganize --vault /path/to/vault --dry-run --json   # stable machine-readable reportdatacron eval --questions examples/eval-questions.example.yaml --vault /path/to/vaultdatacron eval --questions local/golden.yaml --vault /path/to/vault --save-baselinedatacron eval --questions local/golden.yaml --vault /path/to/vault --compare --jsondatacron mcp serve --vault /path/to/vaultdatacron mcp install --client claude-desktop --vault /path/to/vault  # Claude Desktop onlydatacron unregister --client all --scope both --vault /path/to/vaultdatacron protocol uninstall --client all

Current limitations

• Lexical search only: no vector search or embeddings.
• No autonomous agent: the MCP client orchestrates.
• No concurrent multi-machine writes.
• Client detection in datacron setup is best-effort (a config directory or a binary on the PATH); an install in a non-standard location may be missed and can then be configured by hand.

Documentation

Full index: docs/en/index.md | Index français.

To get started:

• Installation and configuration guide
• Use Datacron with Ollama
• Frequently asked questions
• User guide
• Offline library and note consolidation
• Daily memory, people, and commitments
• Essential tool profile

Technical references:

• Vault conventions (SPEC)
• Vault organization
• Runbooks
• Read and reorganize note sections
• Architecture and public surface
• Security boundary
• Integrity scrubber
• Operational health and durability
• Freshness contract

Development

CI runs the invariants and the entire regression suite on Linux/Python 3.12 for changes limited to the READMEs, CHANGELOG, and Markdown pages under docs/fr/ or docs/en/. All other changes retain the six Linux/Windows and Python 3.11-3.13 combinations. Publications require the full matrix, as do empty or unverifiable diffs. ShellCheck, the dependency audit, and the required Quality gate remain active in both paths. The first push of a new branch also uses the full matrix because no previous comparison point is available.

python -m pip install -e ".[dev]"ruff check .ruff format --check .mypypytest

License

Copyright 2026 Julien Bombled.

Licensed under the Apache License, Version 2.0.

Reliable writes and quality gates