I build and harden AI-assisted software focused on reliability, evidence, and clear failure handling. My work spans Python tools, TypeScript applications, and agent workflows.
Authored changes merged into external projects:
- OpenClaw #187880: fixed a Workboard scheduling rule that could leave ready work waiting when completed cards retained stale executions.
- Rosetta #320: improved the performance of a destructive Git command detector while preserving its existing matching behavior.
- Super Productivity #11273: fixed custom weekday handling in the local REST API for recurring tasks.
Source-linked contribution record · Engineering case studies
- Trajectory Fidelity Bridge — converts supported coding-agent transcripts to a portable trajectory format and reports known information loss. Package · Compatibility
- Validation Ledger — a browser-local workspace for connecting product claims, evidence, counterevidence, and decisions. Demo · Reviewer guide
- WeaveStudio — a local-first visual canvas for building and reviewing workflows and deliverables. Public source is available for evaluation under a proprietary license. Demo · Reviewer guide
- CrewAI Stripe Recovery Fixture — local, simulated tests of recovery after ambiguous provider responses; not a live Stripe integration or an exactly-once guarantee.
AI tools assist with research, coding, tests, and drafting. I direct requirements and verification, review changes, and take responsibility for published work.
Technical writing · ResearchOps Community article · Previous profile and additional evidence (historical snapshot; statuses may have changed)
