03 — OPEN SOURCE ECOSYSTEMAUTHORED & MAINTAINED BY @YAKKSHIT

Developer tooling for AI agents that actually work.

Zero-friction CLI utilities and Model Context Protocol (MCP) servers authored by founder Venkata Sai Yakkshit Reddy Asodi. Eliminates context bloat and accelerates local workflows.

FLAGSHIP CLI // @YAKKSHIT/GRAVITON

context-gen-cli: 78% context graph compression.

Autonomous coding agents fail when fed bloated raw repositories.context-gen-cliparses monorepos into AST dependency graphs, pruning dead imports, type aliases, and unreachable symbol tables before streaming directly into Cursor, Claude Code, and Windsurf via MCP.

npm install -g @yakkshit/graviton
—

Cross-Platform AST Engine: Rust-powered parser traverses 100k+ lines of TypeScript, Python, and C++ in under 420ms.

—

Model Context Protocol (MCP v1.2): Native stdio and SSE streaming transports for agentic IDE integration.

—

Direct Token Pruning: Drops boilerplate comments, mock files, and redundant interface definitions to stay inside LLM context limits.

terminal — graviton --prune --mcp
zsh 5.9
$ graviton analyze --repo ./cedzlabs-core --mcp-server
[1/4] Parsing syntax trees across 418 files (TypeScript, Rust)...
[2/4] Constructing call-graph topology... 1,482 nodes resolved.
[3/4] Executing AST pruning pass (mode: aggressive):
· Raw context size: 124,000 tokens
· Pruned context size: 27,280 tokens
· Token reduction: 77.9%
· Dead symbols excised: 642
[4/4] Starting MCP v1.2 server on stdio... READY.
$ graviton export --output .cursor/rules/context.md
✓ Successfully saved 96,720 tokens per prompt cycle.
02 // LIVE AST TOPOLOGY SIMULATOR

Inspect live token pruning graph.

Drag nodes to inspect dependency hierarchy. Hovering illuminates child leaves and calculates token savings.

Token Reduction
78%
Raw vs Pruned
124k → 27,280 tokens
ACTIVE INSPECTION: cedzlabs-core (monorepo)(root)
Raw: 124,000 tokens·Pruned: 27,280 tokens
DEV ENVIRONMENT // 01

nde: Reproducible Development Environments

Streamlined developer environment tooling for fast-moving engineering teams. Spin up reproducible, containerized local environments configured with pre-warmed compilers, linters, and live GGUF inference backends.

CLI: npm install -g ndeGitHub Repository
AI AUTOMATION // 02

CareerAgent: AI Career Automation

Production AI platform automating resume tailoring, ATS parsing, and portfolio synthesis. Features deep MCP integration, zero hallucination guardrails, and typographic PDF compilation.

resume.cedzlabs.comLaunch CareerAgent