A local workbench for recording, replaying, comparing, and inspecting tool-using agent runs without sending trace data to a hosted service.
-
Updated
Aug 4, 2026 - Python
#
agent-harnesses
Agent harnesses are the runtime scaffolding around AI agents. They usually combine context delivery, tool interfaces, planning state, memory, sandboxes, permissions, evaluation, and observability so agents can complete longer tasks reliably. Agent harnesses are especially common in coding agents, research agents, and multi-agent workflows where repeatability, safety, and traceability matter.
Here are 5 public repositories matching this topic...
Open-source framework for reproducible A/B evaluation of agent skills and instructions
python open-source benchmarking reproducible-research ab-testing ai-agents skill-evaluation prompt-engineering llm-evaluation coding-agents agent-evaluation agent-harnesses instruction-evaluation skivolve
-
Updated
Aug 7, 2026 - Python
Codex skill for designing and scaffolding durable agent harness projects, with a Ralph Loop preset and upgrade-friendly doctrine.
developer-tools scaffolding codex ai-agents prompt-engineering agentic-workflows long-running-agents codex-skill ralph-loop harness-engineering project-presets agent-harnesses
-
Updated
Apr 8, 2026 - Python
Interlinked markdown wiki on agent harnesses, orchestration, formal methods, and adjacent research.
-
Updated
May 31, 2026 - TypeScript
Self-updating intelligence dashboard tracking the AI model and tooling landscape — frontier models, agent harnesses, self-hosting, strategy.
-
Updated
Aug 7, 2026 - Python
- Followers
- 108 followers
- Website
- github.com/topics/agent-harness