Skip to main content
Glama
contradictory-body

cc-sensei

by contradictory-body
OverviewSchemaRelated ServersScoreDiscussions
TypeScript
Local

πŸŽ“ cc-sensei

Inject every ounce of Claude Code's architectural wisdom into your AI's brain

English | δΈ­ζ–‡

MCP Tests Modules Latency License

Sound familiar?

😩 "I've already dug deep into Claude Code's source, but when I ask my AI to help me build an Agent, it has no idea what I'm talking about β€” what I read, it didn't."

😫 "I'd love to build a knowledge base so my AI can learn from Claude Code, but the moment I open it β€” 300K+ lines, 1,250 files β€” both my AI and I are scared off."

😡 "I've barely looked at Claude Code's source, but I still want an Agent just as powerful β€” where do I even begin?"

If any of these hit home β€” this project is built for you.

Related MCP server: ontomics

What it does for you

Agent Architecture Oracle distills Claude Code's 310K lines of source and 1,250 files into 32 core modules with 40K lines of structured analysis, and serves it directly to your AI through the MCP protocol.

It covers every core capability of Claude Code:

πŸ” Agent main loop Β· πŸ› οΈ Tool system & execution pipeline Β· πŸ›‘οΈ Permissions & security sandbox Β· 🌊 Model API & streaming 🧠 Context engineering & compaction Β· πŸ“‚ File / Shell / Git tools Β· πŸ”Œ MCP protocol Β· 🎨 Ink rendering engine Β· 🧩 Skills/plugins πŸ’Ύ Persistent memory Β· ⚑ Prompt cache break detection Β· πŸ€– Sub-agents & task system … (22 core + 10 deep-dive supplements)

It establishes a complete index pipeline: natural language β†’ module β†’ design decision β†’ reusable pattern β†’ corresponding source code:

       You say one thing                  Oracle automatically does five things
  ─────────────────────────         ──────────────────────────────────────────
  "How do you compact long       ──►  β‘  Match natural language to M06 (Context Engineering)
   conversations?"                     β‘‘ Return the 4-tier compaction architecture & 9-section summary prompt
                                       β‘’ List 5 directly-copyable engineering principles
                                       β‘£ Cross-trace the 12 modules touched by prompt cache
                                       β‘€ Pull up services/compact/microCompact.ts source on demand

After plugging it into Claude Desktop / Cursor / Qoder / your own Agent, your AI gains a "Chief Architect of Claude Code" as its consultant β€” whether you're cloning the whole thing or just stealing one design, it tells you exactly where to look, why it works, and how to copy it.

In one night, your AI becomes a world-class Agent architect.

πŸš€ Quick Start (3 steps)

Option A β€” Try instantly via npx (zero install)

npx cc-sensei

This downloads and runs the server directly. Skip to Step 3 to connect your AI.

Option B β€” Clone for full control

Step 1 β€” Clone the project

git clone https://github.com/contradictory-body/cc-sensei.git && cd cc-sensei

Step 2 β€” Install and build

Copy-paste again:

pnpm install && pnpm build

πŸ’‘ Don't have pnpm? Run npm install -g pnpm first. πŸ’‘ This step parses 32 module analyses, generates the knowledge index, and compiles the server. One-time only.

When done, you should see: βœ… module-registry.json (32 modules) and βœ… Build success.

Step 3 β€” Connect your AI

Add this snippet to your MCP client's configuration file (replace the path with the one you just cloned to):

{
  "mcpServers": {
    "cc-sensei": {
      "command": "node",
      "args": ["/your/absolute/path/cc-sensei/dist/server.js"]
    }
  }
}

Don't know where the config file is? Common locations:

Client

Config file path

Claude Desktop (macOS)

~/Library/Application Support/Claude/claude_desktop_config.json

Claude Desktop (Windows)

%APPDATA%\Claude\claude_desktop_config.json

Cursor

Settings β†’ MCP β†’ Add new server

Qoder

Settings β†’ MCP β†’ Edit mcpServers

Restart your client. Done! Your AI just gained 6 new skills.

🎯 Try it out

Just say to your AI:

"Use cc-sensei to tell me how Claude Code optimizes prompt cache."

It will call the tools on its own and return a complete analysis with source-code references. That's it.

Dive deeper

πŸ› οΈ What the 6 tools do

Tool

One-line capability

When to use

list_modules

List all 32 modules and their concerns

"Which modules does Claude Code break down into?"

query_architecture

Natural-language search with 3 depth levels (brief/standard/deep)

"How do they prevent long-conversation context overflow?"

get_module

Drill into one module's specific section (responsibility / architecture / decisions / principles / relations)

"Show me M06's design principles"

trace_concern

Trace one concern across all 32 modules, ranked by hit count

"Which modules touch 'prompt cache'?"

search_patterns

Bulk-extract reusable patterns with built-in size limiting

"Give me every cache-related design I can copy"

get_source_code

Read Claude Code source directly (with line numbers, ranges, and directory listings)

"Show me services/api/claude.ts:1412-1456"

Every tool comes with:

  • βœ… Path-prefix tolerance β€” both src/... and bare paths are accepted; no more copy-paste failures

  • βœ… Result rate-limiting β€” 800 chars per section and 12 sections by default, so your AI's context window stays alive

  • βœ… Path-traversal protection β€” attacks like ../../../etc/passwd are rejected outright

  • βœ… Graceful degradation β€” bad arguments return friendly hints instead of crashes

🎬 Real scenario: "I want to add context compaction to my Agent"

You:    "How do you compact conversation history to avoid hitting the token limit?"

Oracle (query_architecture, 103ms):
  β†’ Hit M06: Context Engineering
  β†’ Returns the 4-tier compaction architecture comparison table
    (microCompact / sessionMemoryCompact / autoCompact / reactiveCompact)
  β†’ Trigger conditions, whether LLM is invoked, key file locations β€” all listed

You:    "Show me M06's key design principles"

Oracle (get_module section=principles, 102ms):
  β†’ "Sticky-on Beta Headers β€” performance over simplicity"
    "Single message-level cache_control marker (max 4)"
    "9-section summary structure β€” eval-driven prompt"
  β†’ Each principle includes: where it lives, code snippet, why, how to reuse, the trade-off

You:    "How many modules does 'prompt cache' span?"

Oracle (trace_concern, 118ms):
  β†’ 12 modules ranked by hit count
  β†’ Each with 5 lines of file:line excerpts

You:    "Show me src/services/compact/microCompact.ts"

Oracle (get_source_code, 99ms):
  β†’ Source with line numbers + total line count

A complete "understand β†’ copy β†’ ship" loop, averaging ~110ms.

πŸ“¦ Knowledge scale

πŸ“š 32 module analyses  (40K lines of distilled insight covering 310K lines of source)
β”œβ”€β”€ M01-M22  Core modules
β”‚   M01 Process bootstrap & lifecycle Β· M02 Agent main loop Β· M03 Tool system
β”‚   M04 Permissions & security Β· M05 Model API & streaming Β· M06 Context engineering
β”‚   M07 File/Shell/Git tools Β· M08 MCP protocol Β· M09 LSP integration
β”‚   M10 Bridge/IPC Β· M11 Ink rendering engine Β· M12 Message rendering
β”‚   M13 Input system Β· M14 Sub-agents & tasks Β· M15 Skills/plugins
β”‚   M16 Command system Β· M17 Configuration & settings Β· M18 Telemetry & analytics
β”‚   M19 State management Β· M20 Hotkeys & focus Β· M21 Buddy/voice Β· M22 Testability
└── SUPP-* 10 deep-dive supplements
    AgentSummary Β· SessionMemory Β· autoDream Β· ColorDiff
    extractMemories Β· fileIndex Β· largeFiles Β· speculation
    teamMemorySync Β· yogaLayout

πŸ—‚οΈ Indexes: 390 keywords / 104 concerns / full section-line mappings
πŸ’» Full source code: claude-code-main/src/  available for direct local reads

πŸ—οΈ Engineering philosophy

This project is dogfooding itself β€” built using the methods Claude Code teaches, made to serve Claude.

Claude Code's design philosophy

How this project applies it

Build-time indexing + zero runtime analysis

Section line ranges, keywords, and concerns are all computed in pnpm build:index; runtime just looks up JSON

Path-traversal gatekeeping

validatePath runs before every file read

Graceful degradation > unhandled crashes

Section not found? Return the "available types" hint

Structured prompt-as-spec

All 32 MODULE_NOTES strictly follow a 10-section template

Rate-limiting by default

search_patterns defaults to 800 chars per section, 12 sections cap, configurable

⚑ Performance

p50 = 105ms   p95 = 118ms   max = 125ms
Average response size: 7K chars
Fully local, 0 network dependency

βœ… Tests

pnpm exec tsx scripts/test-suite.ts          # 30 baseline functional tests
pnpm exec tsx scripts/regression-3fixes.ts   # 11 UX-fix regression tests
pnpm exec tsx scripts/e2e-comprehensive.ts   # 24 real-scenario E2E tests

Test suite

Cases

Result

Baseline functionality (6 tools Γ— multiple branches)

30

βœ… 30/30

UX-fix regression

11

βœ… 11/11

Real user-scenario E2E (3 personas)

24

βœ… 24/24

Total

65

βœ… 65/65

πŸ“ Project structure

cc-sensei/
β”œβ”€β”€ src/
β”‚   β”œβ”€β”€ server.ts                 # MCP server entry
β”‚   β”œβ”€β”€ retrieval/engine.ts       # Retrieval engine
β”‚   β”œβ”€β”€ source-reader.ts          # Source reading + path safety
β”‚   β”œβ”€β”€ taxonomy.ts               # 32-module mapping table
β”‚   └── tools/                    # 6 MCP tools
β”œβ”€β”€ scripts/
β”‚   β”œβ”€β”€ build-index.ts            # Parses MODULE_NOTES into indexes
β”‚   β”œβ”€β”€ test-suite.ts
β”‚   β”œβ”€β”€ regression-3fixes.ts
β”‚   └── e2e-comprehensive.ts
β”œβ”€β”€ knowledge/                    # Build artifacts
β”‚   β”œβ”€β”€ module-registry.json
β”‚   β”œβ”€β”€ keyword-index.json
β”‚   └── concern-map.json
β”œβ”€β”€ claude-code-main/             # Claude Code source + 32 module analyses
β”‚   β”œβ”€β”€ src/                      #   ← Source
β”‚   └── MODULE_NOTES/             #   ← Analyses
└── dist/server.js                # Bundle output

βš™οΈ Advanced configuration

Env var

Purpose

Default

CC_SOURCE_ROOT

Claude Code source root

<project>/claude-code-main/src

MODULE_NOTES_ROOT

Module-analysis directory

<project>/claude-code-main/MODULE_NOTES

Point these to anywhere else and Oracle becomes a knowledge server for any project β€” as long as that project has analyses written to the same MODULE_NOTES template.

Roadmap

  • Incremental indexing (only re-parse changed MODULE_NOTES)

  • More section types (performance / security / observability)

  • HTTP / SSE transport (in addition to stdio)

  • Multi-codebase support (mount several projects at once)

License

MIT

Can't read 310K lines of source? Let 40K lines of distilled analysis read it for you, and let your AI copy from it.

If this project helped you, drop a ⭐ so more Agent developers can find it.

Install Server
A
license - permissive license
A
quality
A
maintenance

How are these scores calculated?

Maintenance

–Maintainers
–Response time
–Release cycle
1Releases (12mo)
Commit activity

Resources

Tools

Latest Blog Posts

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/contradictory-body/cc-sensei'

If you have feedback or need assistance with the MCP directory API, please join our Discord server