@pcmstack/mcp is a Model Context Protocol server that lets AI assistants such as Claude Desktop, ChatGPT and Gemini query your Pro Cycling Manager (PCM) databases. Ask about a cyclist's ratings, browse a team's roster, run SQL against the database, or generate a race startlist, all in plain language.

[!IMPORTANT] This server never modifies your existing files. PCM stores its data as binary .cdb files; each call re-reads the .cdb from disk and loads it into an in-memory SQLite database. Every read tool leaves the source untouched. The write tools, pcm_update_database and pcm_update_cyclist_ratings, serialize their changes to a new .cdb file (outputPath) and refuse to overwrite the input, so keep your original as a backup.

Saves and databases

A .cdb file is a database, in Cyanide's own binary format. Your saves are .cdb databases the game writes as you play, but so are the official releases shipped with the game and the community updates you can download, and the tools here work on any of them.

That is why one tool speaks of saves and the rest speak of databases:

• pcm_list_saves finds your saves, across every PCM edition installed on the machine.
• Every other tool takes a databasePath, the path to any .cdb, whether it came from pcm_list_saves or from somewhere else entirely.

Features

• Zero setup: run it with a single npx command, or install a .mcpb bundle with no terminal at all.
• Save discovery: auto-detect your saves on Windows, or point at any .cdb file directly.
• Rich queries: search cyclists and teams, inspect rosters with full per-terrain ratings, and read player info.
• Raw SQL: run guarded, read-only SELECT queries against any table.
• Guarded edits: apply a single INSERT/UPDATE/DELETE, or edit a cyclist's ratings directly, and write the result to a new .cdb, never touching the original.
• Startlist export: generate a PCM-ready startlist XML from a set of teams and rosters.
• Safe by design: read tools are auto-approved by MCP clients; the write tools write only to a separate output file and never overwrite an existing one.

Getting started

Prerequisites

• Node.js 22 or later (not required for the .mcpb bundle install)
• A Pro Cycling Manager database (a .cdb file: a save, an official release or a community update)

Install

Download the latest pcmstack.mcpb from the Releases page and open it with Claude for macOS or Windows. An installation dialog appears, no terminal required.

[!NOTE] This method does not auto-update. To get a newer version, download and re-install the latest .mcpb from the Releases page.

Add the following to your client's MCP configuration file (claude_desktop_config.json, the ChatGPT MCP config, or the Gemini CLI settings file):

{
  "mcpServers": {
    "pcmstack": {
      "command": "npx",
      "args": ["-y", "@pcmstack/mcp"]
    }
  }
}

Once configured, restart your client and ask it something like "list my PCM saves" or "show me the roster of my team".

Platform support

PCM only ships on Windows, where saves live under:

%APPDATA%/Pro Cycling Manager <year>/Cloud/<profile>/

Auto-discovery via pcm_list_saves is therefore Windows only. On macOS/Linux the saves live inside a Wine/Proton prefix that can't be reliably located, so pass an absolute .cdb path directly to pcm_validate_database instead.

Available tools

All tools are prefixed with pcm_, and every one except pcm_list_saves takes an absolute databasePath. Every tool except pcm_update_database and pcm_update_cyclist_ratings is read-only, so clients like Claude Desktop can approve them automatically without a confirmation prompt. The two write tools never overwrite the source or any existing file; they can only create a new .cdb.

| Tool | Description | | ------------------------------ | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ | | pcm_list_saves | Discover your PCM saves on this machine by scanning the Pro Cycling Manager <year>/Cloud folders under %APPDATA%, across every installed edition (Windows only). Returns each save's absolute path (pass it as databasePath to the other tools), plus file name, last modified date and size (newest first). | | pcm_validate_database | Validate that an absolute path points to an existing .cdb database and return its metadata. Stateless: keep the returned path in conversation context to pass to later tools. | | pcm_list_tables | List every table inside a .cdb database, with its ID and name, plus the total table count. | | pcm_get_table_schema | Inspect a single table by name. Returns its columns (name, SQL type, NOT NULL and primary key flags) and its row count. Use pcm_list_tables first to discover available table names. | | pcm_get_player_info | Get the active human player and their team. Returns the player login plus team details (name, resolved division name, resolved country name, evaluation and manager). | | pcm_search_cyclist | Search for a cyclist by first name and/or last name (case-insensitive partial match). Returns up to 10 matches with all ratings (plain, mountain, medium mountain, downhilling, cobble, time trial, prologue, sprint, acceleration, endurance, resistance, recuperation, hill, baroudeur, current ability) and the resolved country name; a truncated flag signals when more matches exist. mediumMountain and currentAbility are null on databases that pre-date those columns. | | pcm_get_team_roster | List a team's roster (defaults to the active player's team when teamId is omitted). Joins DYN_cyclist with its active DYN_contract_cyclist and STA_type_rider; per rider returns name, country, age (derived from birth date and the current game date), rider type, overall ability, contract end year, wage, market value and all per-terrain ability ratings. Ordered by overall ability, highest first. Errors if teamId does not exist. | | pcm_search_team | Search for a team by name (case-insensitive partial match against both the full name and short name). Returns up to 10 matches with the resolved division name, country name, evaluation and general manager; a truncated flag signals when more matches exist. | | pcm_query_database | Run a read-only SQL query (SELECT / WITH … SELECT only) against any table. Write/DDL statements are rejected. Results are capped (default 100, max 1000 rows). | | pcm_update_database | Apply a single INSERT/UPDATE/DELETE statement and write the modified database to a new .cdb at outputPath. The source is never overwritten (outputPath must differ from databasePath); SELECT, schema changes (DROP/CREATE/ALTER) and stacked statements are rejected. Returns the written path and the number of rows changed. | | pcm_update_cyclist_ratings | Change one or more ability ratings of a cyclist (by IDcyclist) and write the modified database to a new .cdb at outputPath. Takes a ratings object where each field is optional (plain, mountain, mediumMountain, downhilling, cobble, timeTrial, prologue, sprint, acceleration, endurance, resistance, recuperation, hill, baroudeur; 50–85); only the fields provided are changed. Returns the written path and the cyclist's full ratings after the update. Setting mediumMountain is rejected on databases that pre-date that column. | | pcm_generate_startlist_xml | Generate a PCM startlist XML document from a list of teams and their cyclist rosters. Looks up the race by IDrace to derive the output file name from STA_race.gene_sz_filename (e.g. c0_almeria.xml), and returns both the file name and the XML as text. Team and cyclist IDs map to DYN_team.IDteam / DYN_cyclist.IDcyclist (look them up with pcm_search_cyclist or pcm_query_database). |

How it works

Tools are stateless: there is no "current database" held by the server. Every tool takes an absolute databasePath, re-validates it, and re-reads the .cdb from disk into a fresh in-memory SQLite database (via cdb-converter + sql.js) for each call. The file on disk is never mutated: read tools only ever read it, and the write tools (pcm_update_database, pcm_update_cyclist_ratings) write their changes to a separate output .cdb. A typical flow is:

1. pcm_list_saves (Windows) to find your saves, or pcm_validate_database with an explicit path to point at any .cdb.
2. pcm_search_cyclist, pcm_get_team_roster, pcm_query_database, … to explore it.
3. pcm_generate_startlist_xml to produce a startlist file for a race, or pcm_update_cyclist_ratings / pcm_update_database to write an edited copy.

Development

Clone the repo and install dependencies with npm install, then:

npm run build     # bundle src/ -> dist/ with tsup (ESM output)
npm test          # run the vitest suite once
npm run test:watch  # vitest in watch mode
npm run coverage  # vitest with v8 coverage
npm run lint      # biome lint --write . (autofixes)
npm run format    # biome format --write .
npm run pack      # produce dist/pcmstack.mcpb

To debug the server interactively with the MCP Inspector:

npm run build && npm run inspector