repositories
Open-source projects by Vikas Udasi — AI agents, cloud systems, and experimental tools.
Profile
/vikasudasi — AI & Cloud Architecture, Engineering Leadership
Repositories
| Repository | Description |
|---|---|
| skill-vault | Self-hostable, semantic skill registry for AI agents over a single MCP endpoint — search skills by task, retrieve what you need on demand, verify what you trust via sha256 + ed25519 signing |
| eu-act-check | CLI compliance scanner for EU AI Act Article 50 — scans files for C2PA manifests, EXIF metadata, and AI disclosure text patterns with pass/warn/fail reports |
| no-slop | CLI that scans your codebase for AI slop patterns and generates constraint files (SKILL.md, .cursorrules, CLAUDE.md, .clinerules) for Claude Code, Cursor, Codex, and Gemini CLI |
| mcp-app-suite | MCP Apps development suite: playground to preview interactive HTML UIs, scaffolder to scaffold new projects, and demo server with 3 sample apps |
| mcp-apps-render | CLI that renders MCP Apps interactive UI payloads (dashboards, forms, visualizations) directly in the terminal — no browser, no iframe, no HTML |
| agent-creds-scanner | CLI that scans AI agent config files (CLAUDE.md, .cursorrules, MCP configs) for hardcoded credentials before they leak |
| cache-smith | CLI that analyzes your prompts for KV cache potential and simulates semantic caching with local embeddings — benchmark savings before buying the gateway |
| llm-context-fmt | CLI that reveals how format constraints (JSON, XML, Markdown, plain text) systematically bias LLM outputs — based on July 2026 ArXiv research |
| readme2skill | CLI that reads your README and docs to auto-generate agent skill definitions (SKILL.md) — compatible with Claude Code, Codex, Cursor, and Gemini CLI |
| vuln-seeker | CLI code security scanner — detects vulnerabilities with 25+ regex patterns and optional AI-powered triage, with JSON/Markdown reports and CI mode |
| doc-inject-guard | CLI that scans documents (DOCX, PDF, MD, HTML) for hidden prompt injection payloads — 5 detection modules, risk scoring, CI integration |
| budget-forge | CLI that finds the optimal thinking budget for reasoning LLMs — test prompts at multiple budget levels and generate cost-vs-quality tradeoff curves |
| model-cards | CLI reference tool for the open-weight AI model explosion — search, view, and compare model cards from Hugging Face in your terminal |
| slop-detect | Privacy-preserving CLI that detects AI-written text using 5 local heuristics — fully offline, no API calls |
| prompt-riser | CLI that runs prompts through multiple reasoning strategies and visualizes the reasoning paths as a tree |
| skill-audit | CLI that audits, validates, and inventories agent skills (SKILL.md files) across any ecosystem |
| pr-classify | CLI that detects AI-generated PRs and generates audit reports for human reviewers |
| prompt-snapshot | Zero-dependency CLI that snapshots project context for AI agent prompts |
| vikasudasi.github.io | Personal portfolio & open-source AI agent projects |
| my-agent-project | Multi-agent task management system with CLI, MCP, and dashboard |
| my-agent | Personal macOS deep agent for terminal automation |
| my-agent-voice | Voice layer for my-agent — speak to your terminal agent |
| context-aware-rag | 100% local, privacy-preserving RAG with bilingual OCR |
| ai_coding | Agent-friendly project structure for AI coding agents |
| company-os | AI-native operating system for company workflows |
| AI_Travel_agent_Streamlit | AI travel agent built with Streamlit |
| console_web_search | Web search from your terminal — no browser needed |
| restaurant-bot | Restaurant discovery and recommendation bot |
| hello-world | Hello World |
| mycollection | Personal collection |