hermes-oh-my-pi-mcp
A tiny Hermes-friendly bridge that exposes oh-my-pi's CLI as MCP tools (prompt and doctor) for code generation and environment verification.
README
hermes-oh-my-pi-mcp
A tiny Hermes-friendly bridge for oh-my-pi / omp.
This repository now includes both the working MCP bridge and the full integration explanation for GitHub readers.
What this repo provides
It exposes two MCP tools:
prompt— runompin one-shot mode and return the outputdoctor— verifyompandbunare available locally
Why this exists
Hermes works well with MCP, plugins, and custom tools. oh-my-pi already provides a strong CLI/runtime, so this bridge lets Hermes users call it without reimplementing the agent stack.
The recommended operating model is:
User → Hermes → MCP bridge → bun → omp → result
Hermes vs MCP vs Skill
- Hermes: decides what to do
- MCP bridge: actually calls
omp - omp: performs the coding task
- Skill: teaches Hermes when to prefer
ompand how to fall back when needed
In other words, the skill is helpful but optional; the MCP bridge is the execution layer.
Detailed integration guide
See the full explanation here:
That guide covers:
- call structure
- why MCP is the right wrapper
- when a skill helps
- example Hermes configuration
- troubleshooting and usage policy
Release notes
Use this file to see what changed in each published version.
Requirements
- Node.js 20+
ompinstalled and onPATHbuninstalled and onPATHbecauseompis Bun-based
Install oh-my-pi
npm install -g @oh-my-pi/pi-coding-agent
If the omp binary fails with bun: No such file or directory, install Bun 1.3.14+ and retry.
Local development
npm install
npm test
npm run doctor
Run as an MCP server
npx hermes-oh-my-pi-mcp
Or from source:
node ./src/index.js
Hermes config example
Add this to Hermes MCP config:
{
"mcpServers": {
"oh-my-pi": {
"command": "npx",
"args": ["-y", "hermes-oh-my-pi-mcp"]
}
}
}
If you want to point Hermes at a local checkout instead:
{
"mcpServers": {
"oh-my-pi": {
"command": "node",
"args": ["/absolute/path/to/hermes-oh-my-pi-mcp/src/index.js"]
}
}
}
Verified local setup
On this machine, Hermes is actually connected to oh-my-pi and the MCP server is enabled.
The working registration looks like this:
hermes mcp add oh-my-pi \
--command node \
--args /home/ubuntu/oh-my-pi-hermes-mcp/src/index.js \
--env OMP_BIN=/home/ubuntu/.local/share/mise/installs/node/24.16.0/lib/node_modules/@oh-my-pi/pi-coding-agent/dist/cli.js \
--env BUN_BIN=/home/ubuntu/.bun/bin/bun
That setup was verified with:
hermes mcp test oh-my-pi
If your machine uses different paths, keep the same shape but replace the absolute values with your own OMP_BIN and BUN_BIN locations.
Suggested skill policy
If you also want Hermes to prefer omp for coding requests, add a small skill that says:
- Code generation / refactoring / analysis requests should consider
ompfirst. - If the environment looks broken, run
doctorfirst. - If
ompoutput is insufficient, Hermes should supplement it with its own reasoning and tools.
That skill is not required for the bridge to work; it just makes the usage pattern consistent.
Notes
- The bridge runs
ompin a one-shot JSON-friendly mode and returns the text response. - Set
OMP_BINif your binary is not justomp. - Set
BUN_BINif Bun is installed somewhere custom. - The long-form architecture and usage notes live in
docs/hermes-omp-integration.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.
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.
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.
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.
Neon Database
MCP server for interacting with Neon Management API and databases
E2B
Using MCP to run code via e2b.
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.