World Model Mcp
by SaravananJaichandarcommunityv0.8.1
4
2
An MCP server for Claude Code that builds a temporal knowledge graph of a codebase from session events, user corrections, and test outcomes, then uses it to validate edits, learn constraints, and reduce regressions. It stores local project data in SQLite and integrates with hooks and external linters for pre-edit enforcement.
From the registry: Temporal knowledge graph for codebases with constraint enforcement at the edit boundary.
Install & configure
Prerequisites
- required
- requiredInstall the world-model-mcp packageInstall the MCP server from PyPI using pip.
https://pypi.org/project/world-model-mcp/$ pip install world-model-mcp - requiredRun project setupInitialize the server in your project and auto-seed the knowledge graph. This creates project files such as .mcp.json and .claude settings/hooks.
$ python -m world_model_server.cli setup
Install in your client
Paste into the AI you're chatting with — it will figure out the rest
Please install the `world-model-mcp` MCP server into my current AI client (that's you).
Required prerequisites (do these first if not already done):
- **Install Python 3.11+** — Python 3.11 or newer is required. (https://www.python.org/downloads/)
- **Install the world-model-mcp package** — Install the MCP server from PyPI using pip. Run: `pip install world-model-mcp` (https://pypi.org/project/world-model-mcp/)
- **Run project setup** — Initialize the server in your project and auto-seed the knowledge graph. This creates project files such as .mcp.json and .claude settings/hooks. Run: `python -m world_model_server.cli setup`
Canonical MCP server config (stdio transport):
- command: `world-model`
- args: []
- optional environment variables:
- `WORLD_MODEL_DB_PATH`: Custom path for the local world model database directory. Defaults to ./.claude/world-model/ if unset. (example: `<your-db-path>`)
- `ANTHROPIC_API_KEY`: Anthropic API key to enable LLM-powered extraction and constraint inference. The server works without it using regex-based fallback behavior. (example: `<your-anthropic-api-key>`)
- `WORLD_MODEL_EXTRACTION_MODEL`: Anthropic model name used for extraction tasks. (example: `<your-extraction-model>`)
- `WORLD_MODEL_REASONING_MODEL`: Anthropic model name used for reasoning tasks. (example: `<your-reasoning-model>`)
- `WORLD_MODEL_DEBUG`: Enable debug mode. (example: `<your-debug-setting>`)
Note: The README’s preferred portable path is the PyPI install. The project setup command generates project-local configuration automatically, but the canonical MCP command shape is the stdio server executable `world-model`. The server is local-first and can run without ANTHROPIC_API_KEY; that key only enables optional LLM features.
Add this MCP server to my current client's config in the correct format for you. If you need secrets or credentials I haven't provided, ASK me — do not invent values or leave raw placeholders. After adding it, tell me how to verify the server is connected.Note: The README’s preferred portable path is the PyPI install. The project setup command generates project-local configuration automatically, but the canonical MCP command shape is the stdio server executable `world-model`. The server is local-first and can run without ANTHROPIC_API_KEY; that key only enables optional LLM features.
Tools (8)
query_fact
Check whether APIs or functions exist before using them.
record_event
Capture development activity with reasoning chains.
validate_change
Validate a proposed change against learned constraints and linters before execution.
get_constraints
Retrieve project-specific rules for a file or path.
record_correction
Learn constraints from user corrections and edits.
get_related_bugs
Assess regression risk and find related bugs for a file or change.
seed_project
Scan a codebase and populate the knowledge graph with entities and relationships.
ingest_pr_reviews
Import GitHub PR review comments and turn feedback into constraints.
Environment variables
WORLD_MODEL_DB_PATHrequiredDatabase location for the world model storage.ANTHROPIC_API_KEYrequiredEnables optional LLM-based extraction and reasoning.Prerequisites
- • Python 3.11+
Packages
world-model-mcp
pypiv0.8.1transport: stdio
Details
- Language
- Python
- License
- MIT
- Released
- Jun 17, 2026
- Last commit
- May 30, 2026
- Registry name
- io.github.SaravananJaichandar/world-model-mcp
Related Servers
Labelhead Artist Momentum
Trending hip-hop artist momentum scores across four cultural dimensions.
73.6k
PostHog MCP Server
Official PostHog MCP Server for product analytics, feature flags, experiments, and more.
35.5k
Openmetadata Mcp
Official OpenMetadata MCP: governed context and business semantics for AI assistants and agents.
14.5k