The open-source app everyone uses to manage agents at work
TypeScript acpxagent-orchestrationagent-skillsagentic-aiai-agentsai-engineeringai-governanceai-toolsautonomous-agentsclaudeclaude-codeclaude-skillscodexhermesllmmcpmulti-agentopenclawproject-managementworkflow-automation
Runner SDK and lab documentation
These documents cover the package-local browser SDK, React components, standalone demos, scenario explorer, issue-thread surface, and live-session development tools. They do not enable a production provider or change Paperclip's runtime selection.
Tutorials
- Local runner
- Live console protocol server
- Live console
- SDK console
- Standalone adapter
- Scenario explorer
- Scenario chat
- Issue-thread lab
- Clean-room Codex chat
Reference
- Runner eval scoring slice
- Deterministic runner workflow evals
- Evals integration contract
- Local runner and supervision
- Durable transport and recovery
- Live console protocol server
- Live console
- Browser and React SDK
- Standalone adapter
- Scenario explorer
- Scenario chat
- Mock control plane
- Live runnerd/Codex loop
- Issue-thread UI
- Clean-room chat
- Issue-thread UX contract
- Scenario explorer UX
- Scenario chat UX
- Live-console interaction map
- Live-console component decisions
- SDK component decisions
- Standalone adapter boundary
- UI library compatibility note
The SDK and labs are additive. Existing direct adapters continue to use their existing composer, transcript, interaction, and finalization paths.