Index of personal software projects and experiments — AI agents, web apps, dev tools, and platform integrations. Each project now lives in its own repo; this README is the map.
AI agents register and discover each other — a social graph for the agentic web.
- Tech Stack: Next.js, Vercel, TypeScript, AI Agents
- Status: Shipping
Real-time visibility and coordination layer for multi-agent AI systems.
- Tech Stack: OpenClaw, Python, SQLite, WebSocket
- Status: Active
Real-time dashboard for the Indian Avengers AI org — 25+ agents, cron jobs, security telemetry, GitHub activity.
- Tech Stack: Python, Chart.js, OpenClaw
- Status: Active
AI-powered PM assistant — turns scattered requirements into structured product specs.
- Tech Stack: Claude, Next.js, TypeScript
- Status: Active
macOS menu bar app for token tracking, budget enforcement, and local LLM monitoring.
- Tech Stack: Swift, macOS, Ollama
- Status: Production
InterviewAgent (showcase)
Public-facing project home: polished README, daily features/ and quality/ delivery journal.
- Tech Stack: docs-only — source lives in the implementation repo below
- Status: Production
interview-agent (implementation)
Multi-agent pipeline for job applications — resume optimization, cover letters, auto-submission at 500+/day.
- Tech Stack: Python, Streamlit, Playwright, Supabase, Claude, MCP
- Status: Production
AI coaching platform with web and iOS apps — learning plans, progress tracking, subscription management.
- Tech Stack: Python, Flask (web), Swift (iOS), Kivy (mobile)
- Status: Active
Modular agent framework with planning, execution, and memory (Python, YAML workflows).
- Tech Stack: Python, YAML
- Status: Experimental
Enterprise edge security platform for Indian mid-market — real-time threat detection, AI anomaly detection, edge computing security dashboard.
- Tech Stack: Next.js 15, TypeScript, Tailwind CSS, Framer Motion
- Status: Active
AI-powered phishing simulation and instant coaching MVP for SMB teams.
- Tech Stack: Node.js, Express
- Status: MVP
- This repo used to be a monorepo with each project as a subfolder. In 2026-04 the three substantive projects (
CoachAI,InterviewAgent,ai_agent_project) were extracted into their own repos viagit subtree splitwith history preserved. Several exploratory stubs (ClaudeCode,ClaudeCode-v2,WikiRevamp,windsurf-demo,Project Starlink,Azure DevOps MCP,Ghostty) were removed. - Some of the linked repos are private — links will 404 for anyone without access.
- Sensitive config files (
.env,secrets.env) are not in version control. - Agent activity is logged to
autonomy-log.jsonlper workspace. - Monitor all active agents via Mission Control.
MIT — see also CODE_OF_CONDUCT.md, CONTRIBUTING.md, SECURITY.md.
- Functional: Add project health scorecards with milestone progress, blockers, and confidence levels
- Non-functional: Add integration tests for project CRUD and permissions with CI gating
- Functional: Add project dependency graph view with critical path highlighting
- Non-functional: Optimize project list query performance with indexed filters