Search

GuidingHand SDK: train AI agents like new hires. Agents, files, procedures, conversations, evals, sessions, tasks, recordings and webhooks.

OpenCode plugin (V1 + V2): runs every subagent asynchronously in the background, a TUI sidebar monitor with token stats, opt-in background-subagent control tools (children/result/cancel/send), opt-in event-driven progress reports from a child's todowrite

SLICC's agent core on pi durable: an agent worker, session storage, remoting and pi extensions

Beta — run labelling for the Niyoos proxy: wraps your own OpenAI client so every call carries its run id, step and parent run.

Standalone compute-native agent harness.

Standalone compute-native agent harness.

AI agent that generates and tests Scrape Admin profiles using only the run/validate endpoints.

Conviva tracing SDK for AI agents — a thin wrapper over OpenLLMetry that exports OTLP http/protobuf traces to the Conviva agent gateway, with Conviva's span processors and baggage connector built in.

Standards-oriented security SDK for Agent Nebula and Agent Genesis

Spraay Protocol batch payment tool for BeeAI agents — USDC batch payments on Base via the x402 gateway

Mercury 3 is a diffusion-based language model from Inception, and the lab's most capable model to date. It is suited for agentic workflows and real-time applications where latency compounds, such...

Your X (Twitter) bookmarks aren't a graveyard anymore: your AI agent (Claude Code, Codex, Omnigent or any CLI) reads each one, works out why you saved it, and writes a task you start with one click.