Developer Tooling & AI
Open-source agent tooling and developer workflows
OpenClaw contributions in TypeScript, mostly around agent runtime diagnostics, Gateway behavior, cron/session handling, plugin validation, and CLI reliability.
Queue backlog observability walkthrough
Uses an executable local model instead of prose so every signal claim can be regenerated from the companion repository.
Open the evidenceScope
Specific capabilities
The narrower skills inside this area.
Examples
Where this shows up
Public examples and pages that connect the skill to code, decisions, or delivery work.
Project implementation
There is working implementation behind the example, with source and tests.
- Queue backlog observability walkthrough
Uses an executable local model instead of prose so every signal claim can be regenerated from the companion repository.
Professional context
Where I have used it
My OpenClaw work is a good fit for the part of agent tooling that looks like systems engineering: finding the runtime boundary, making the failure mode visible, adding focused tests, and keeping the fix narrow enough for a large open-source project to review.