Cm codeintell

Skill tody-agent/codymaster/skills/cm-codeintell

Unified code intelligence — routes to Skeleton Index, CodeGraph, Architecture Diagram, or Smart Context Builder based on task shape. Loads deep refs on demand.From its SKILL.md

Install

npx -y skills add tody-agent/codymaster --skill cm-codeintell

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

no licenseNo license file was found in the repository. Code published without one is not open source by default, so using it at work is a question for whoever answers licensing questions where you are.

SKILL.md

2.7 KB, 552 tokens by cl100k_base, as published. Nobody here has run it

Code Intelligence — Structural Understanding for AI Agents

TL;DR

Use to index a codebase for fast structural understanding
Layers: skeleton, code graph, architecture diagram, smart context
Default: start from the lightest useful layer
Next: cm-planning, cm-debugging, or cm-execution

Stop scanning. Start querying. Load only the layer the task actually needs.

When to Use

Understanding a medium or large codebase
Tracing callers, callees, or impact before edits
Building focused context for planning, debugging, or execution
Auto-triggering structural understanding during project setup

Choose a Layer

Need instant codebase orientation with zero extra setup?
└─ YES → Layer 0: Skeleton Index

Need symbol search, callers/callees, or impact analysis?
└─ YES → Layer 1: Code Graph

Need a visual system / module map?
└─ YES → Layer 2: Architecture Diagram

Need a focused context packet for another skill or agent?
└─ YES → Layer 3: Smart Context Builder

Layer	Summary	Load
0	Grep-based skeleton index for instant orientation	`references/layer-0-skeleton.md`
1	AST graph and symbol-query workflow	`references/layer-1-codegraph.md`
2	Mermaid architecture generation	`references/layer-2-architecture.md`
3	Synthesized focused context for downstream work	`references/layer-3-context-builder.md`

Conditional References

Load only the target layer reference for the current task.
Load references/integration-workflows.md only when wiring cm-codeintell into cm-start, cm-planning, cm-debugging, or cm-execution.

Integration

Skill	How `cm-codeintell` helps
`cm-start`	bootstrap structural context early
`cm-planning`	impact analysis and component boundaries
`cm-debugging`	trace callers, callees, and failure paths
`cm-execution`	pre-flight context for focused implementation

Rules

Start from the lightest layer that answers the question.
Prefer Layer 0 when simple orientation is enough.
Regenerate or refresh indexes after major structural changes.
Do not pay for deep graph or context synthesis unless the task needs it.

The Bottom Line

Pick the cheapest layer that answers the question, then load deeper only when needed.

What ships with it: 5 files

5.8 KB alongside SKILL.md

references/

Gives 0 of the 12 instructions most diagrams charts skills give in 552 tokens

Counted across 397 of the 418 authors here whose files we hold, read 2026-09-06

Quote labels containing special charactersin 18 of 397, across 17 files
Choose the right diagram type for the datain 15 of 397, across 13 files
Create a detailed Mermaid diagramin 12 of 397, across 11 files
Validate Mermaid syntax before outputin 11 of 397, across 9 files
Wrap diagrams in fenced Mermaid code blocksin 11 of 397
Read the reference file for the chosen diagram typein 10 of 397, across 6 files
Label every connection descriptivelyin 8 of 397, across 7 files
Use stateDiagram-v2 for state diagramsin 7 of 397
Declare the diagram type on the first linein 7 of 397
Keep one concept per diagramin 7 of 397
Cap diagrams at 15 nodes and split over 20in 7 of 397
Describe the diagram before generating itin 7 of 397

Said here and by no other author read

Start from the lightest layer that answers the question
Prefer Layer 0 for simple orientation
Load the integration reference only when wiring into other skills
Regenerate or refresh indexes after major structural changes

Grouped from the skills themselves: near-identical wordings counted once, and counted by distinct author, so one author publishing three of these counts once. Length counted with cl100k_base; the agent that loads this file may tokenize it differently.