More screenshots
Works with the AI Agents you already use
CodeContext Pro — AST Code Intelligence
Reduce agent token usage by 88% using AST-based surgical code editing and semantic repository mapping.
$29
CodeContext Pro — AST Code Intelligence
Example session with this skill installed
Use AST surgical tools to read the processCheckout function in src/services/paymentService.ts and add a timeout parameter with a default of 5000.
- Read your context and instructions
- Compiled the codecontext pro
I have performed a surgical AST read of the requested symbol and prepared the refactoring. The change adds a timeout parameter to the processCheckout method signature.
Surgical AST Refactor: processCheckout
| Metric | Value |
|---|---|
| File Path | src/services/paymentService.ts |
| Symbol Target | processCheckout |
| Mutation Type | ast_transform (signature_update) |
| Token Savings | 88.2% (380 vs 3,210 tokens) |
| Confirmation Token | ctx_ref_882_x92j_001 |
--- src/services/paymentService.ts
+++ src/services/paymentService.ts
@@ -42,7 +42,7 @@
- public async processCheckout(order: Order): Promise<PaymentResult> {
+ public async processCheckout(order: Order, timeout: number = 5000): Promise<PaymentResult> {
const transaction = await this.gateway.initialize(order);
// ... logic continued
Next steps
- Review the diff above for correct parameter positioning and default value.
- Run the confirmation command:
write_file_surgical(confirm: true, confirmationToken: "ctx_ref_882_x92j_001"). - If the logic is incorrect, run
rollback_fileto clear the pending mutation.
Connects securely to your tools. The creator never sees your data.
About this skill
CodeContext Pro v4.0.0 Enterprise — AI Brain Operating System
"Never feed a 1,000-line file into your attention window when a 15-line symbol is all you need."
CodeContext Pro transforms standard AI coding agents (Claude Code, Cursor, Windsurf, Codex, Copilot) into Lead Staff Software Architects. By replacing brute-force file reads (cat, read_file) and unindexed searches (grep, find) with deterministic Tree-sitter WebAssembly Abstract Syntax Tree (AST) operations, CodeContext Pro cuts token overhead by 75% to 88%, eliminates hallucinated symbol edits, and enforces atomic two-phase mutations with instant rollback.
⚡ Certified Performance Benchmarks (222 Tests Verified)
| Repository Stack | Raw File Tokens | CodeContext AST Tokens | Tokens Saved | Token Reduction |
|---|---|---|---|---|
| Rails Monolith (120 files) | 184,200 | 28,500 | 155,700 | 84.5% |
| Laravel 11 App (95 files) | 142,000 | 24,100 | 117,900 | 83.0% |
| FastAPI / Django (80 files) | 118,500 | 19,800 | 98,700 | 83.3% |
| TypeScript / React SPA (150 files) | 210,000 | 36,400 | 173,600 | 82.7% |
| Single Controller Edit (1,200 lines) | 3,850 | 480 | 3,370 | 87.5% |
🚀 Key Capabilities & Tools
- The Syntax Firewall: Pre-commit Tree-sitter validation prevents broken syntax, unmatched braces, or hallucinated edits from reaching your disk.
- Incremental AST Delta Re-parse: Sub-millisecond syntax caching (< 0.8 ms) enables lightning-fast symbol refactoring without freezing your session.
- Inverted LSP Type Skeleton: Extracts complete interface, class, and method signatures without running heavy 500MB language server daemons.
- Virtual ORM Triad: Extracts ActiveRecord, Laravel Eloquent, and Django model schemas directly from migrations and code — without requiring a live database connection.
- Interactive Architecture Visualizer: Generate self-contained, zero-dependency HTML repository maps with real-time DOM search filtering.
- Two-Phase Atomic Commits: Preview git unified diffs with cryptographic confirmation tokens and instant one-step rollback.
⚡ 60-Second Setup
Claude Code CLI:
claude mcp add code-context -- npx -y mcp-code-context@latest
Cursor & Windsurf:
Add to your MCP settings as command:
npx -y mcp-code-context@latest
Or connect via Streamable HTTP:
npx -y mcp-code-context@latest --http --port=3000
URL: http://127.0.0.1:3000/sse
📦 What's Inside This Package
SKILL.md: Complete cognitive directives, negative trigger guards, and decision tree for AI agents.references/: Benchmarks, decision matrix (AST vs Grep vs I/O), and multi-IDE configuration guides.scripts/: Platform-agnostic installation and verification scripts for Windows, macOS, and Linux.7
Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Skills for LLM Ops.
Requires Node.js environment. Virtual schema extraction limited to Rails, Laravel, and Django. Token savings vary based on file size and symbol density.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before buying a skill.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Money-back guarantee
Doesn't do what its page promises? You get your money back.
Trust & safety
Security scan
Security scan passed 19 days ago
- Software
- Code Context MCP
Node.js 18+Windows, macOS, Linux (Universal)
Permissions
File access is strictly scoped to the active project workspace for AST analysis and two-phase surgical edits. All parsing runs 100% locally via WebAssembly with zero network telemetry or tracking.