Project Directory

Sets non-stopping traps on functions in a live Python AI agent, capturing arguments, locals and return values without ever pausing the process.

Turns a code repository into a guided tour with an architecture graph and user-journey traces, every claim checked against real line numbers.

A terminal tool that fixes Python bugs with Claude, checks each fix by re-running the code, and explains the lesson through a short philosophical monologue.

Connects customer-reported shop issues to an engineering dashboard where Claude investigates logs, metrics and code, proposes a patch and verifies it with tests.

A portal that hands each registered event participant an anonymised, single-use credit link, synced from the event’s guest list.

Scans an uploaded project for leaked secrets, malicious-looking code, corrupted, duplicate and unnecessary files, cleans what it safely can and scans again.

An AI security agent that runs locally, attacks your own application and shows exactly which requests worked and what data it could reach.

"Postman for MCP": connect to any MCP server (SSE, Streamable HTTP, stdio), test tools in a schema-generated playground, and track P50/P95/P99 latency, schema drift and security issues, with a CLI that emits JSON/JUnit for CI/CD.