AI coding work records and validation tool
A personal tool linking AI work across machines to requirements, issues, execution and validation records. I set goals and change boundaries, then checked the implementation, tests and deployments performed by coding agents.
Screens
Reading requests and results in context
I connected work states by machine, project and conversation to requirements and issues. The tool helps check whether work satisfies its original goal, scope and acceptance criteria.
Human decisions and agent execution
I define features, priorities and acceptance criteria; agents implement, test and deploy. Changes that fail required checks are corrected and checked again. Optional independent reviews remain distinct from mandatory validation.
Conditions for updates and recovery
I required verification that an update’s processes and files are separate from ongoing conversations before applying it. I checked rollback and redeployment records for the web and API, along with preservation of conversations and original records.
Architecture
