claude-historian-mcp

An MCP server for searching your Claude Code conversation history: past solutions, errors, plans and file change history.

  • 180 ★ on the repository
  • Last commit
  • MIT
MCP server

Extended access

Why this level

  • The server reads local logs of all Claude Code dialogs, including code and secrets inside them.
  • Search results land in the model context.
All reasons and checks

vvkmnn/claude-historian-mcp

Install

In your terminal, with SkillFoxx CLI

npx skillfoxx add mcp/claude-historian-mcp

Detects the agents on your machine, shows the access level and pins the version.

Other ways to install

Access and checks

How we review

Extended accessThe tool runs code or makes network calls and only reads external data.

  • The server reads local logs of all Claude Code dialogs, including code and secrets inside them.
  • Search results land in the model context.

This is third-party code. Review the repository files before installing.

What it does

claude-historian-mcp gives Claude access to your conversation history. It runs locally with no external dependencies or databases, only search algorithms. The server has two tools, `search` and `inspect`. `search` covers eleven scopes: conversations, files, errors, plans, config, tasks, sessions, tools, similar queries and project memories. Parameters set the timeframe, project and level of detail. For example `search query="ENOENT" scope="errors"` finds past errors with their fixes, and `search filepath="package.json" scope="files"` shows the operation history of a file. `inspect` builds a short session summary by ID: files, tools, decisions and outcomes. A separate skill teaches Claude to check history on its own before web searches, on errors and at session start. A plugin from the claude-emporium marketplace adds hooks and a `/historian-search` command.

Who it is for. Claude Code users with a long work history who want to find past solutions and errors without searching the web again.

Good fit when

  • The same `ENOENT` error appeared a month ago and you need to find how it was fixed.
  • You want to see which files changed in a session and which decisions were made.
  • Before a web search it is worth checking whether the agent already answered a similar question.
  • A plan discussed with Claude in a neighboring project has to be recalled.

Not a fit when

  • Your history lives in another client: the server reads only Claude Code logs.
  • You need semantic memory with embeddings: here search is algorithmic with no vector database.
  • Logs were cleared or disabled and there is nothing to search.

Example request

Search my history for how we fixed the docker auth error and show the files that changed then.

MCP

Transport
stdio
Authentication
not required

Limitations

The server reads local Claude Code conversation logs, so it is useful only with accumulated history and can surface content of past dialogs, including secrets in them. Search is algorithmic with no embeddings, and precision depends on query wording and parameters. The skill improves discovery, though the MCP also works without it. The project was renamed from claude-historian to claude-historian-mcp, so old configs need updating. The hooks plugin requires the server to be installed first.

How to disable. Run `claude mcp remove claude-historian-mcp`, remove the skill with `npx skills remove` or by deleting its directory, and the plugin with `/plugin uninstall`.

FAQ

How do I connect the server?

Run `claude mcp add claude-historian-mcp -- npx claude-historian-mcp`. For Cursor and Windsurf add the same `npx claude-historian-mcp` to `mcp.json`.

Which tools exist?

Two: `search` for eleven scopes and `inspect` for a session summary by ID or the value `latest`.

Is a database needed?

No. The server runs locally with no external dependencies and no separate databases.

Why install the skill?

The skill makes Claude check history on its own before web searches, on errors and at session start. Install it with `npx skills add Vvkmnn/claude-historian-mcp --skill claude-historian --global`.

What does the plugin add?

Hooks before WebSearch, before plan mode, before subagents and after Bash errors, plus a `/historian-search` command. It installs from the `Vvkmnn/claude-emporium` marketplace.

Discussion

No questions or reviews yet.

A SQLite archive of Claude Code sessions with an MCP server, a terminal UI and a web UI for finding past conversations.

CLIExtended access64Repository stars

claude-history searches and opens past Claude Code, Codex and Pi conversations in the terminal and gives Claude a history skill

CLIExtended access542Repository stars

Local memory for Claude Code: a single binary indexes past conversations and injects relevant context into new sessions.

MCP serverBroad access230Repository stars

A local web editor for Claude Code config: shows project sizes, deletes old chat histories and manages MCP servers

CLIExtended access266Repository stars

An MCP server and CLI that find in Claude Code and Codex sessions why a file became what it is: who asked the agent what, when and what changed

MCP serverExtended access776Repository stars

Minutes

minutes

Official

A local-first MCP server for meeting memory: on-device recording and transcription into Markdown, searchable from Claude Code, Codex, Cursor and any MCP client

MCP serverExtended access1.6KRepository stars

README in short

The server works over stdio, installs nothing locally and searches Claude Code logs. Three install paths are described, the skill and the plugin, then the `search` and `inspect` tools with call examples. It closes with search methodology, comparison with alternatives and a developer guide.

README badge

Report a problem