# icm-graph

> Use this tool when you need to optimize AI coding agent interactions, reducing token usage by 70-90% through context bundles and semantic memory. The icm-graph CLI tool takes in coding requests and outputs optimized code responses, utilizing BM25, symbol-level navigation, and output filters to improve recall. It is ideal for use cases involving large-scale coding projects, such as git integrations, where efficient AI agent communication is crucial.

Canonical page: https://skillsregistry.net/skills/ncmonx-icm-graph  
JSON: https://api.skillsregistry.net/v1/skills/ncmonx-icm-graph

## Description

Token-efficient CLI for AI coding assistants. 70-98% cheaper Claude API. 40 MCP tools, 913/913 tests, Apache-2.0.

## Trust

- **Trust score (0–1):** 0.65
- **Verification tier:** scanned

## Facts

- **Version:** 1.0.0
- **Skill type:** atomic
- **Execution layer:** container
- **Runtime environment:** vm
- **Category:** other
- **Updated:** 2026-06-16

## Source

- **Source listing:** [GitHub](https://github.com/ncmonx/icm-graph)

## Use it

Resolve this record through the SkillsRegistry MCP server (no auth, read-only):

```
claude mcp add --transport http --scope user skillsregistry https://api.skillsregistry.net/mcp
```

```json
{
  "jsonrpc": "2.0",
  "id": 1,
  "method": "tools/call",
  "params": {
    "name": "get_skill",
    "arguments": {
      "slug": "ncmonx-icm-graph"
    }
  }
}
```

REST: `GET https://api.skillsregistry.net/v1/skills/ncmonx-icm-graph` · pull for local use: `GET https://api.skillsregistry.net/v1/skills/ncmonx-icm-graph/pull`

---
SkillsRegistry indexes agent skills from public registries and GitHub. Skills we have analysed are scanned with Circle-IR and scored on six dimensions; each listing states its scan coverage. More: https://skillsregistry.net/llms.txt