cm-codeintell

Name: cm-codeintell
Author: tody-agent

47stars23forks

How to install cm-codeintell?

tody-agent/codymaster/cm-codeintell

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

Installs into the current project.

Summary

Unified code intelligence — Skeleton Index (zero-dep, <4s) + AST knowledge graph (CodeGraph) + architecture diagrams (Mermaid) + smart context builder. Pre-indexes code structure so AI agents understand any codebase instantly. 95% token compression for onboarding. 30% fewer token

Files · 1

View on GitHub

SKILL.md

1	# Code Intelligence — Structural Understanding for AI Agents
2
3	## TL;DR
4	- Use to index codebase for fast structural understanding
5	- Outputs: skeleton index, AST graph, Mermaid diagrams
6	- Token saving: ~95% for onboarding, ~30% for deep analysis
7	- Next: cm-planning or cm-debugging
8
9	> Stop scanning. Start querying. Skeleton Index (<4s, zero deps) + AST graph + architecture diagrams = instant code understanding.
10	> Inspired by [CodeGraph](https://github.com/colbymchenry/codegraph) + [GitDiagram](https://github.com/ahmedkhaleel2004/gitdiagram).
11	> TRIZ-optimized: 10 inventive principles applied.
12
13	## When to Use
14
15	ALWAYS for medium-to-large projects. This is infrastructure, not an action skill.
16
17	- Auto-triggered by: `cm-start` Step 0.7 (project init) — ALWAYS runs Layer 0
18	- Manually triggered for: "understand this codebase", "what calls X?", "what breaks if I change Y?"
19	- Skip when: NEVER — Layer 0 (Skeleton) works on any project size
20
21	### Detection Thresholds (Auto-Trigger)
22
23	```
24	TRIGGER if ANY of these are true:
25	→ Project has >50 source files
26	→ User wants to refactor or re-code an existing project
27	→ User says "understand the codebase" / "what does this do?"
28	→ cm-execution encounters >3 grep/glob calls for one task
29	→ cm-debugging needs callers/callees to trace a bug
30	```
31
32	---
33
34	## Architecture: 4 Layers
35
36	```
37	┌──────────────────────────────────────────────────────────────────────────┐
38	│ cm-codeintell │
39	├──────────────────┬──────────────────┬──────────────────┬────────────────┤
40	│ LAYER 0 │ LAYER 1 │ LAYER 2 │ LAYER 3 │
41	│ Skeleton Index │ Code Graph │ Architecture │ Smart Context │
42	│ (Instant) │ (Structure) │ Diagram (Visual)│ (Synthesis) │
43	├──────────────────┼──────────────────┼──────────────────┼────────────────┤
44	│ grep/find/awk │ tree-sitter AST │ File tree + LLM │ All layers + │
45	│ → skeleton.md │ → SQLite graph │ → Mermaid.js │ qmd → focused │
46	│ (~5K tokens) │ → MCP server │ → .cm/ storage │ context packet │
47	├──────────────────┼──────────────────┼──────────────────┼────────────────┤
48	│ ZERO deps │ codegraph_* │ Auto-generated │ Feeds: exec, │
49	│ <4 seconds │ MCP tools │ at project init │ plan, debug │
50	│ ANY project size │ 50+ files │ 20+ files │ All consumers │
51	└──────────────────┴──────────────────┴──────────────────┴────────────────┘
52	```
53
54	### TRIZ Principles Applied
55
56	\| # \| Principle \| How Applied \|
57	\|---\|-----------\|-------------\|
58	\| #1 Segmentation \| 4 independent layers — each usable alone \|
59	\| #2 Taking Out \| Extract only signatures, discard function bodies \|
60	\| #5 Merging \| CodeGraph + GitDiagram + Skeleton → one unified skill \|
61	\| #10 Prior Action \| Pre-index at project init, not at query time \|
62	\| #13 Inversion \| Code summarizes ITSELF to agent (push, not pull) \|
63	\| #15 Dynamicity \| Adaptive: skeleton (<20) vs graph (>50) vs full (>200) \|
64	\| #25 Self-Service \| Auto-detect project size → auto-select intelligence level \|
65	\| #28 Mechanics Substitution \| Replace file reading (slow) with pattern matching (fast) \|
66	\| #35 Parameter Changes \| Unit: file content → function signature → 95% compression \|
67	\| #40 Composite \| One skill = skeleton + graph + diagrams + context builder \|
68
69	---
70
71	## Layer 0: Skeleton Index (Instant — Zero Dependencies)
72
73	> Purpose: Lightning-fast grep-based extraction of function signatures, class definitions, exports, and module boundaries. Produces a compact `.cm/skeleton.md` that gives the agent instant understanding of any codebase.
74
75	### How It Works
76
77	```
78	1. SCAN → find all source files (14 languages supported)
79	2. EXTRACT → grep for function/class/export signatures only
80	3. GROUP → organize by directory (module boundaries)
81	4. CAP → limit per-dir (15 files) + total (600 lines)
82	5. OUTPUT → .cm/skeleton.md (~5K tokens for 600-file project)
83	```
84
85	### Usage
86
87	```bash
88	# Run from project root
89	bash scripts/index-codebase.sh
90
91	# Custom paths
92	bash scripts/index-codebase.sh /path/to/project /path/to/output.md
93	```
94
95	### What It Extracts (Per Language)
96
97	\| Language \| Patterns Extracted \|
98	\|----------\|-------------------\|
99	\| TypeScript/JavaScript \| `export`, `function`, `class`, `interface`, `type`, `enum`, `const =`, routes \|
100	\| Python \| `def`, `async def`, `class`, `@app.route`, `from...import` \|
101	\| Go \| `func`, `type...struct`, `type...interface`, `package` \|
102	\| Rust \| `pub fn`, `struct`, `enum`, `impl`, `trait`, `mod` \|
103	\| Java/Kotlin \|

People who liked this also viewed

6 picks

Type

cm-codeintell

47stars23forks

View on GitHub

How to install cm-codeintell?

tody-agent/codymaster/cm-codeintell

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

Installs into the current project.

Summary

Files · 1

View on GitHub

SKILL.md

1	# Code Intelligence — Structural Understanding for AI Agents
2
3	## TL;DR
4	- Use to index codebase for fast structural understanding
5	- Outputs: skeleton index, AST graph, Mermaid diagrams
6	- Token saving: ~95% for onboarding, ~30% for deep analysis
7	- Next: cm-planning or cm-debugging
8
9	> Stop scanning. Start querying. Skeleton Index (<4s, zero deps) + AST graph + architecture diagrams = instant code understanding.
10	> Inspired by [CodeGraph](https://github.com/colbymchenry/codegraph) + [GitDiagram](https://github.com/ahmedkhaleel2004/gitdiagram).
11	> TRIZ-optimized: 10 inventive principles applied.
12
13	## When to Use
14
15	ALWAYS for medium-to-large projects. This is infrastructure, not an action skill.
16
17	- Auto-triggered by: `cm-start` Step 0.7 (project init) — ALWAYS runs Layer 0
18	- Manually triggered for: "understand this codebase", "what calls X?", "what breaks if I change Y?"
19	- Skip when: NEVER — Layer 0 (Skeleton) works on any project size
20
21	### Detection Thresholds (Auto-Trigger)
22
23	```
24	TRIGGER if ANY of these are true:
25	→ Project has >50 source files
26	→ User wants to refactor or re-code an existing project
27	→ User says "understand the codebase" / "what does this do?"
28	→ cm-execution encounters >3 grep/glob calls for one task
29	→ cm-debugging needs callers/callees to trace a bug
30	```
31
32	---
33
34	## Architecture: 4 Layers
35
36	```
37	┌──────────────────────────────────────────────────────────────────────────┐
38	│ cm-codeintell │
39	├──────────────────┬──────────────────┬──────────────────┬────────────────┤
40	│ LAYER 0 │ LAYER 1 │ LAYER 2 │ LAYER 3 │
41	│ Skeleton Index │ Code Graph │ Architecture │ Smart Context │
42	│ (Instant) │ (Structure) │ Diagram (Visual)│ (Synthesis) │
43	├──────────────────┼──────────────────┼──────────────────┼────────────────┤
44	│ grep/find/awk │ tree-sitter AST │ File tree + LLM │ All layers + │
45	│ → skeleton.md │ → SQLite graph │ → Mermaid.js │ qmd → focused │
46	│ (~5K tokens) │ → MCP server │ → .cm/ storage │ context packet │
47	├──────────────────┼──────────────────┼──────────────────┼────────────────┤
48	│ ZERO deps │ codegraph_* │ Auto-generated │ Feeds: exec, │
49	│ <4 seconds │ MCP tools │ at project init │ plan, debug │
50	│ ANY project size │ 50+ files │ 20+ files │ All consumers │
51	└──────────────────┴──────────────────┴──────────────────┴────────────────┘
52	```
53
54	### TRIZ Principles Applied
55
56	\| # \| Principle \| How Applied \|
57	\|---\|-----------\|-------------\|
58	\| #1 Segmentation \| 4 independent layers — each usable alone \|
59	\| #2 Taking Out \| Extract only signatures, discard function bodies \|
60	\| #5 Merging \| CodeGraph + GitDiagram + Skeleton → one unified skill \|
61	\| #10 Prior Action \| Pre-index at project init, not at query time \|
62	\| #13 Inversion \| Code summarizes ITSELF to agent (push, not pull) \|
63	\| #15 Dynamicity \| Adaptive: skeleton (<20) vs graph (>50) vs full (>200) \|
64	\| #25 Self-Service \| Auto-detect project size → auto-select intelligence level \|
65	\| #28 Mechanics Substitution \| Replace file reading (slow) with pattern matching (fast) \|
66	\| #35 Parameter Changes \| Unit: file content → function signature → 95% compression \|
67	\| #40 Composite \| One skill = skeleton + graph + diagrams + context builder \|
68
69	---
70
71	## Layer 0: Skeleton Index (Instant — Zero Dependencies)
72
73	> Purpose: Lightning-fast grep-based extraction of function signatures, class definitions, exports, and module boundaries. Produces a compact `.cm/skeleton.md` that gives the agent instant understanding of any codebase.
74
75	### How It Works
76
77	```
78	1. SCAN → find all source files (14 languages supported)
79	2. EXTRACT → grep for function/class/export signatures only
80	3. GROUP → organize by directory (module boundaries)
81	4. CAP → limit per-dir (15 files) + total (600 lines)
82	5. OUTPUT → .cm/skeleton.md (~5K tokens for 600-file project)
83	```
84
85	### Usage
86
87	```bash
88	# Run from project root
89	bash scripts/index-codebase.sh
90
91	# Custom paths
92	bash scripts/index-codebase.sh /path/to/project /path/to/output.md
93	```
94
95	### What It Extracts (Per Language)
96
97	\| Language \| Patterns Extracted \|
98	\|----------\|-------------------\|
99	\| TypeScript/JavaScript \| `export`, `function`, `class`, `interface`, `type`, `enum`, `const =`, routes \|
100	\| Python \| `def`, `async def`, `class`, `@app.route`, `from...import` \|
101	\| Go \| `func`, `type...struct`, `type...interface`, `package` \|
102	\| Rust \| `pub fn`, `struct`, `enum`, `impl`, `trait`, `mod` \|
103	\| Java/Kotlin \|

People who liked this also viewed

6 picks

Type