diff --git a/docs/claude/error-book.md b/docs/claude/error-book.md new file mode 100644 index 000000000..a5d7b9439 --- /dev/null +++ b/docs/claude/error-book.md @@ -0,0 +1,12 @@ +# Claude's error book (MeltMind) — read before every ruling + +Started 2026-11-19 at Yanqi's request ("各自犯错和误解的东西都该自己想办法记住…但凡多看看我发过的消息都会好很多"). + +## Recurring errors → check that prevents them +1. **Scope drift from the goal** (made "SOTA on a public 3D benchmark" a main line → ShapeNet car pressure). Check: every ruling names its link to "usable fill-time prediction on unseen parts" in one line; anything without that link is context, not a main line. +2. **Overstating** ("route closed" for Adviser; "systematic, not seed noise"; "packing does not affect fill" before logs; "needs only OK"; "organisation complete" before cleanup/answers). Check: before posting, label each claim observed / inferred / untested; use "not established" rather than "closed/impossible". +3. **Reading Yanqi's messages too narrowly.** Check: re-read her full original wording (not TRON's summary) before ruling; list each of her questions and answer each one; a launch is not a deliverable. +4. **Process heavier than the risk** (5+ review rounds for a 4 GPU-h run). Check: reviews scale with cost and risk; guard edge cases that cannot cause runaway, target reads or wrong science are known limitations. +5. **Unverified technical advice** (suggested a browser User-Agent; Horsea 403 is route-level, not UA). Check: test a suggestion from my own route before recommending it, and say when the target environment may differ. +6. **Adoption/gate criteria mixing validity and attribution** (1b timing item). Check: gates test validity only; attribution is a separate field. +7. **Roles:** TRON relays and wakes; Grok gathers information (research, social media), not heavy review; Claude decides.