Sources show what the advice rests on
Read three inspections, the validation record, and the video evidence. Each separates inspected code from reported results and work not run. An account association does not establish authorship.
Inspected code puts advice in context
Read the T3 Code and Melee inspection to separate advice, inspected code and measured results. Later code does not establish the setup at filming time.
- A completion contract specifies the real output, how complete its inputs must be and the outcome its consumer sees.
- Recurring mistakes need checks at existing lint, type, test and build boundaries that reject bad changes and accept valid ones.
- Explicit ownership separates agent writes and test state, contains provider quirks in adapters and checks remote access to artifacts.
- Project decisions stay beside relevant code, while evidence distinguishes local runs, remote CI, author reports and untested hardware.
Shared language connects the inspected workflows
Read the course video project inspection for shared language and recovering interrupted work. Account association alone does not establish authorship.
- Shared vocabulary, lasting decisions, bounded modules and callable agent APIs inform the lesson on speaking a codebase's language.
- Output identity, resumable work and recovery from malformed results inform the lesson on artifacts that survive interruptions.
- The report recommends making the intended outcome easy to request and hard to misreport.
Different checks reveal different output failures
Read the public-project inspection for several ways to check useful output. Tests were inspected, not run locally.
- The report discusses layered validation to test public output, reproduce new failures and preserve meaningful baselines.
- Before-and-after measurements compare the same work to establish whether a change improved the result.
- Existing lessons on lint, adapters and native loading provide supporting references for the skills.