|
26 | 26 |
|
27 | 27 | ## 5. Mistake Reflection Protocol |
28 | 28 |
|
| 29 | +> [!NOTE] for Users: |
| 30 | +> `docs/ai/ai_learnings.md` is a local-only file (excluded from git) meant to be |
| 31 | +> read by the developer to improve AI assistant behavior on this project. Use its |
| 32 | +> findings to improve the GEMINI.md setup. |
| 33 | +
|
29 | 34 | When you realise you have made a mistake — whether caught by the user, |
30 | 35 | by a tool, or by your own reasoning — you MUST: |
31 | 36 |
|
32 | 37 | 1. **Acknowledge the mistake explicitly** and explain what went wrong. |
33 | | -2. **Reflect on the root cause**: was it a missing check, a false |
34 | | - assumption, skipped verification, or a gap in the workflow? |
35 | | -3. **Immediately append a new entry to @./docs/ai/ai_learnings.md** |
36 | | - following the format defined in that file. This is not optional and |
37 | | - does not require user confirmation. Do it before continuing. Update user |
38 | | - about the changes to the workflow in the current chat. |
| 38 | +2. **Reflect on the root cause**: was it a missing check, a false assumption, skipped verification, or a gap in the workflow? |
| 39 | +3. **Immediately append a new entry to `docs/ai/ai_learnings.md`** — this is not optional and does not require user confirmation. Do it before continuing, then update the user about the workflow change. |
| 40 | + |
| 41 | + **Entry format:** |
| 42 | + - **Mistake**: What went wrong. |
| 43 | + - **Root cause**: Why it happened. |
| 44 | + - **Rule**: The concrete rule added to prevent recurrence. |
39 | 45 |
|
40 | 46 | The goal is to treat every mistake as a signal that the workflow is |
41 | 47 | incomplete, and to improve it in place so the same mistake cannot |
|
0 commit comments