Agent Skills Hub

reverse-engineering-api

Reverse engineer web APIs by capturing browser traffic (HAR files) and generating production-ready Python API clients for automation and data extraction.

sqs

AWS SQS skill for managing message queues, decoupling microservices, configuring dead-letter queues, handling visibility timeouts, and implementing FIFO ordering.

mls

Unified local ML inference server for ASR, TTS, Translation, Image Generation, and Vision on Apple Silicon, powered by MLX.

log-focus-debug

Focus debug skill for DashPlayer: isolates log chains, injects temporary focus markers ([FOCUS:token]), and ensures clean removal of debug artifacts after task completion.

compatibility-testing

Orchestrate cross-browser, cross-device, and responsive design testing using cloud providers like BrowserStack and Playwright to ensure consistent user experiences.

text-to-speech

Expert Kokoro TTS implementation skill for real-time, secure, and offline voice synthesis in JARVIS-style assistants. Features streaming output, prosody control, and performance-optimized audio generation.

milan-news-briefing

Automate your daily Milan news digest with this Python-based briefing tool. Supports weather, strikes, world/AI/Italian news, and event scraping, featuring deduplication, RSS/API pipeline management, and AI-agent ready scheduling.

Observability

Real-time observability dashboard for PAI multi-agent activity, featuring live WebSocket streaming, session tracing, and agent workflow debugging.

railway-database

Provision and manage Railway database services (Postgres, Redis, MySQL, MongoDB) with automated configuration and environment wiring.

evaluating-code-models

Evaluate code generation models using BigCode Evaluation Harness. Benchmarks include HumanEval, MBPP, and MultiPL-E with pass@k metrics for multi-language coding models.

ray-skill

Sends debugging data, logs, and visual output to the Ray desktop application via its local API for real-time developer feedback.

guard

Epistemic safety analysis for JSON data in prompts to prevent LLM hallucinations and reasoning errors when handling incomplete or large-scale datasets.