Search

Portable, framework-agnostic Node agent core for Ready, Set, Cloud: a Strands-TS assistant with DynamoDB-backed conversation snapshots and pluggable cross-session memory.

Provider-neutral, local-first operations layer for token frugality and safer AI-agent work.

An auditable local workspace for Pi: pinned upstream source, curated extensions, bounded MCP tooling, and a reproducible external runtime.

Production-ready agentic AI framework in Go; build scalable reasoning agents with native tool calling, memory & observability. Anthropic · OpenAI · Gemini · Groq

Satellite-native crop yield intelligence: Spatio-Temporal Transformer + live Sentinel-2/ERA5 via Google Earth Engine, FastAPI backend, and an LLM tool-calling agent (Groq/Llama 3.3) with guardrails and automated eval. Research prototype — synthetic training data, documented known issues.

Profiler and optimization layer for AI agent execution graphs

Agent-ready command-line interface for Lexware Office — contacts, invoices, and the receivables/AR-aging the API refuses to total for you.

Take your agents live: cron and file-watch dispatch, safety wrappers, and operations for standard agent definitions

PyPI wrapper for TenantGuard, a tenant-isolation security-audit CLI for self-hosted multi-tenant AI-agent platforms. Downloads and runs the official prebuilt Go binary from GitHub Releases.

WACP — turn any backend into an AI-agent-friendly system via /ai/* endpoints.

SF CLI plugin: McCabe cyclomatic complexity and Agent LOC analysis for AgentScript (.agent) bundles.

Standalone evidence-first debugging CLI for coding agents

Run a prompt through a LangChain (system + human) chain over Gemini on Vertex AI; optional LangSmith tracing.

SDK for interacting with Cortex Code CLI programmatically

Reconstructs why a file or code region exists from its introducing commit, later edits, co-changes, and intent clues

Vei AI Agents - Modern CLI AI Coding Agent using Gemini 3.5 Flash

LangChain integration for SERPdive, the AI Search API and Tavily alternative: answer-ready web content for agents, same speed, 20.2% fewer tokens, higher answer quality (60.7% of decided duels) on a public benchmark

Flight-recorder style work framework for strong AI coding agents — records follow work, they don't lead it. ✈️

Research Prototype: 12-Agent Personalized Learning System. Multi-agent collaboration, EventBus, Memory Layer, 6 resource types, self-improvement loop.

Define LLM-powered steps as typed async Python functions.