Build, run and scale AI agents like API and microservices - observable,auditable and identity-aware from day one.

A reusable Agent Skill that turns research briefs into narrated vertical explainer videos with a pre-render review gate.

A Codex-first agent skill that turns spatial data into polished, verified interactive maps and report-ready figures.

Autonomously improve a real artifact (code, training recipe, agent harness, data pipeline, prompt) against an objective and an evaluator, using Hypothesis Tree Refinement (HTR) from the Arbor paper. Use this whenever someone wants to iteratively optimize something over many experiments without overfitting — e.g. "get my model's eval score up", "improve this agent/harness", "tune this pipeline", "beat the baseline on this benchmark", "run a search over approaches and keep the best", "do an MLE-bench / Kaggle-style optimization", or any long-horizon "make this artifact better and don't just memorize the dev set" task. Trigger it even when the user doesn't say "Arbor" or "hypothesis tree" but describes repeated experiment-and-evaluate loops, branching exploration of competing ideas, or worries about a dev/test gap. Runs Claude itself as the coordinator with subagent executors in isolated git worktrees; for the standalone `arbor` CLI tool see references/arbor-upstream.md.

Open-source, local-first desktop Agent Runtime for turning business goals into executable, reusable, and auditable workflows.

A Codex skill for auditable, local-first Chinese book-review short-video workflows with dual styles and provenance tracking.

Open, evidence-bounded GEO and SEO agent skill hub with reusable workflows for discovery, diagnosis, content, and measurement, packaged for AI agent runtimes.

AI-native stock screening engine with full-market discovery, LLM ranking, risk-aware scoring, and auditable evaluation. AI选股

Lightweight Windows tray app that monitors your Claude rate limits in real time - portable single EXE, zero configuration, fully auditable

An installable DeepSeek Harness plugin that packages a nine-stage foreign-trade SDR workflow with lead dedup, human approval, and audit capabilities.

SPD V2 Quality. Discover, deeply qualify, submit, and audit truthful product listings across Windows, macOS, and Linux-capable environments on relevant product, software, startup, app, and AI-tool directories with audience-value screening, SEO quality gates, action-level authorization, privacy controls, and evidence-backed records. Use when listing quality, referral value, governance, durability, and auditable compliance matter more than submission volume. Do not use for bulk backlink creation, ranking manipulation, guest posts, community promotion, mass email, paid link acquisition, or automatic submission to low-quality directories.

Lightweight TDD skill suite for AI coding agents enforcing test-first workflow.