A

agent

An AI that reads your metrics, logs and traces at 3 a.m. so it can tell the half-awake on-call engineer exactly which of their own deploys started the fire, all on your own hardware so nobody else sees the shame.

Your cloud SI is smart, but it can't touch your computer. Homing Pigeon fixes that: tell your SI to complete a task and it will do it on your machine (Linux/macOS/FreeBSD/Windows), and the results return so the SI can iterate: run the tests, read the failure, fix, repeat. Think 'GitHub self-hosted runner, but the control plane is email.' Python 3.9+, standard library only, no dependencies. Jobs are HMAC-SHA-384-signed JSON carried by email (protocol v2); a local Oracle model reviews what comes back before your SI sees it. By design it executes signed jobs with the runner account's privileges: run it on a machine you can afford to lose.

The skills, docs and report template that Korvus reads before every incident and every release test, kept in a public repo where anyone can watch it get ready for the next disaster.

The core Swytchcode Execution Kernel for AI Agents & Developers.

A living memory system for the AI agent(s) that assist your daily work. Supported by ChatGPT, Claude, V0 and many others. Get started in 5 minutes with our managed service, or clone the repo and host completely free.

Personal PI agent harness.

Agent-shape testing harness that measures how an LLM-driven agent uses a tool's CLI, scored by an LLM judge.

Agentic coding CLI — any OpenAI-compatible API, any model, any shell.

Agent autonome de détection d'anomalies de marché (LangGraph + MCP + RAG)

Part of my collection of skills, created with Crush with different model families (Qwen, Deepseek) targeting localhosted models like QwenX 27b, and other mid sized models

Scan agents for reliability issues in GitLab CI/CD

https://github.com/trustabl/trustabl

Slack as yourself, for an agent: one Go binary that is both an MCP server (stdio) and a slack CLI, authenticated with a user OAuth token. Includes the full procedure and approved request text for a GitLab team member to get their own app approved (docs/setup.md), and an AGENTS.md for the agent doing it.

A simple agent that does meal suggestions and produces shopping lists based on meal suggestions

A comprehensive climate agent for weather forecasts, geocoding, and pollen data.

https://demsking.gitlab.io/oremi-ojo

MCP server for controlling Zigbee2MQTT smart home devices.

https://demsking.gitlab.io/oremi-isakoso/

An MCP server for discovering public radio stations and stream URLs through Radio Browser.

https://demsking.gitlab.io/oremi-rediyo/

Composable Agent primitives for Kit applications

🧠 Open-source persistent memory engine for AI agents and LLMs with semantic search, automatic deduplication, and intelligent context retrieval.

Personal skills and agent configuration

Agentics provides a complete AI infrastructure ecosystem covering every aspect of building modern AI-powered applications

from language model inference across 27+ models, to voice processing with TTS/STT/VAD/EOT, semantic embeddings, secure networking with tunnels and VPN, real-time financial data streaming, and comprehensive developer tools including CLI and mobile apps.

The modular architecture allows businesses to use individual components or the entire suite, with production-ready deployment, monitoring, and South African payment integration. Whether you're building a voice assistant, an AI-powered application, or need secure infrastructure for remote access and real-time data, Agentics provides the complete toolkit