Odoo OSI MCP
Enables coding agents to search, inspect, and recommend open-source Odoo/OCA modules before writing custom code, with source-backed evidence and version/migration guidance.
README
Odoo Open Source Intelligence
Odoo Open Source Intelligence, or Odoo OSI, is an enterprise-grade platform for discovering, evaluating, comparing, and recommending open-source Odoo modules before teams write custom code.
The AI-facing interface is Odoo OSI MCP, an unofficial MCP server that lets coding agents search and inspect indexed OCA and trusted open-source Odoo repositories.
Odoo OSI is not affiliated with, endorsed by, or sponsored by the Odoo Community Association (OCA) or Odoo S.A. OCA and Odoo names are used only to identify public open-source repositories and ecosystem references.
Current Build Stage
This repository now has the first working backend and MCP vertical slice:
- FastAPI API for health, repositories, modules, search, code search, and solution finding
- PostgreSQL schema and Alembic migration for repositories, branches, modules, dependencies, source files, and symbols
- GitHub/OCA ingestion for repositories, Odoo version branches, manifests, metadata, dependencies, and selected source files
- Odoo-aware Python, XML, CSV security, and README extraction
- requirement-to-module search with version, license, manifest, and indexed document signals
- source-backed recommendation evidence
- cross-version recommendation candidates with migration/backport guidance
- live GitHub/OCA fallback for likely modules that are not indexed yet
- MCP server with search, solution, code, module, and dependency tools
- persisted indexing job ledger surfaced through
/indexing/status - local PostgreSQL and Redis configuration
- linting and test coverage
Quick Start
Follow these steps to run Odoo OSI locally from a fresh clone.
1. Prerequisites
Install:
- Python 3.11 or newer
- Docker and Docker Compose
- Git
- A GitHub personal access token for practical indexing and fallback search
2. Clone The Repository
git clone https://github.com/krishnakantjoshi/odoo-osi-mcp.git
cd odoo-osi-mcp
3. Create Local Configuration
cp .env.example .env
Open .env and set your own GitHub token:
ODOO_OSI_GITHUB_TOKEN=YOUR_GITHUB_TOKEN
.env is ignored by git. Do not commit real tokens.
4. Start PostgreSQL And Redis
docker compose up -d postgres redis
5. Create A Python Environment
python -m venv .venv
source .venv/bin/activate
pip install -e ".[dev,mcp]"
On Windows PowerShell, activate the virtual environment with:
.\.venv\Scripts\Activate.ps1
6. Run Database Migrations
alembic upgrade head
7. Start The API Server
uvicorn odoo_osi.api.app:create_app --factory --reload
Check the API:
curl http://127.0.0.1:8000/health
curl http://127.0.0.1:8000/indexing/status
8. Discover And Index A Small Sample
In another terminal, activate the virtual environment again:
source .venv/bin/activate
Discover module manifests from one OCA repository:
odoo-osi discover-oca \
--repository purchase-workflow \
--odoo-version 18.0 \
--branch-limit 1 \
--module-limit 20 \
--persist
Index source evidence for a small set of modules:
odoo-osi index-source \
--repository purchase-workflow \
--odoo-version 18.0 \
--module-limit 10 \
--file-limit 30
Check coverage:
odoo-osi coverage
curl http://127.0.0.1:8000/indexing/coverage
9. Search For Modules
curl -X POST http://127.0.0.1:8000/solutions/find \
-H "Content-Type: application/json" \
-d '{"requirement":"multi level purchase approval","odoo_version":"18.0","limit":5}'
10. Run The MCP Server
For a local MCP client, use stdio:
odoo-osi run-mcp
Most MCP clients use a config like this:
{
"mcpServers": {
"odoo-osi": {
"command": "/absolute/path/to/odoo-osi-mcp/.venv/bin/odoo-osi",
"args": ["run-mcp"],
"cwd": "/absolute/path/to/odoo-osi-mcp",
"env": {
"ODOO_OSI_GITHUB_TOKEN": "REPLACE_WITH_YOUR_OWN_GITHUB_TOKEN"
}
}
}
}
The same example is in mcp.json.example. If your MCP client does not support
env, put the token in your local .env instead.
11. Run Tests
python -m ruff check .
python -m pytest
Example Prompts
After configuring the MCP server in your AI coding tool, try these prompts.
Check Index Coverage
Use the odoo-osi MCP server.
Before searching for modules, call get_coverage_report and summarize:
1. how many modules are indexed
2. how many modules have source evidence
3. whether results should be treated as complete or partial
4. what indexing step I should run next
Find An Existing Module Before Custom Development
Use the odoo-osi MCP server.
I am developing on Odoo 18 Community.
Requirement:
I need a multi-level purchase approval workflow.
Before writing custom code, check whether an existing open-source Odoo/OCA module satisfies
this requirement.
Return exact-version matches first, then older-version migration candidates, then discovered
but not indexed candidates. Include module name, repository, license warning, evidence level,
version status, migration effort, source evidence, README evidence, security-rule evidence,
and indexing guidance.
Do not generate custom code until I review the MCP result.
Inspect A Candidate Module
Use the odoo-osi MCP server.
Inspect the module purchase_request from OCA/purchase-workflow for Odoo 18.0.
Summarize:
1. manifest metadata
2. dependencies
3. license warnings
4. models and inherited models
5. views, actions, and menus
6. security access rules
7. README evidence
8. whether it is safe to use as implementation reference
Search For A Code Pattern
Use the odoo-osi MCP server.
Search indexed OCA source for modules that extend purchase.order in Odoo 18.0.
Return the repository, module, file path, symbol type, model name, inherited model, and any
useful XML IDs.
Prepare A Development Brief
Use the odoo-osi MCP result as the primary reference.
Target:
- Odoo version: 18.0
- Edition: Community
- Organization/developer: Your Org
- Technical module prefix: your_org
- New module technical name: your_org_purchase_approval
Produce a development brief, branding brief, exact file plan, and minimal test plan.
Do not write code until the brief and file plan are approved.
For full prompt templates, see AI Development Brief Guide.
Troubleshooting
- Database connection fails: run
docker compose up -d postgres redis, then rerunalembic upgrade head. - GitHub rate limits: set
ODOO_OSI_GITHUB_TOKENin.env. - MCP client cannot find
odoo-osi: use the absolute path to.venv/bin/odoo-osiin your MCP config. - Search returns few results: run discovery and source indexing first. A fresh database has no indexed modules.
Local Credentials
There are two different GitHub credentials you may use:
- GitHub personal access token: used by Odoo OSI to call the GitHub API for indexing and fallback search.
- SSH key: used by git itself when you push or pull repositories over SSH.
GitHub API Token
Create a fine-grained personal access token in GitHub:
- Open GitHub.
- Go to Settings.
- Open Developer settings.
- Open Personal access tokens.
- Choose Fine-grained tokens.
- Generate a new token.
- Use the minimum access needed. For public OCA indexing, read-only public repository access is enough.
Put that token in your local .env file:
ODOO_OSI_GITHUB_TOKEN=YOUR_GITHUB_TOKEN
.env is ignored by git. If your MCP client uses a private mcp.json, that file is ignored too.
Do not put real tokens in mcp.json.example, docs, tests, or committed source.
GitHub docs:
Git SSH Key
For git push/pull access over SSH, keep your SSH key outside the repo, usually in ~/.ssh, and
add the public key to your Git hosting provider.
On macOS/Linux:
ssh-keygen -t ed25519 -C "your_email@example.com"
eval "$(ssh-agent -s)"
ssh-add ~/.ssh/id_ed25519
cat ~/.ssh/id_ed25519.pub
Copy the public key output and add it in GitHub under Settings > SSH and GPG keys > New SSH key. Never copy or commit the private key file.
GitHub docs:
API Endpoints
The health endpoint:
GET /health
Key working endpoints:
GET /repositories
GET /modules
POST /search
POST /search/code
POST /solutions/find
Roadmap
See ROADMAP.md.
Ingestion
See docs/ingestion.md for GitHub token setup and indexing commands.
MCP
See docs/mcp.md for the Odoo OSI MCP server commands and tools.
Project Status And Notices
- This project is unofficial; see NOTICE.md.
- Contributions are welcome; see CONTRIBUTING.md.
- Security reporting and credential guidance are in SECURITY.md.
- Privacy and hosted deployment notes are in PRIVACY.md.
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.