$npx -y skills add toroleapinc/claude-brain --skill brain-conflictsReview and resolve unresolved brain merge conflicts.
| 1 | The user wants to resolve pending brain merge conflicts. |
| 2 | |
| 3 | ## Steps |
| 4 | |
| 5 | 1. Read the conflicts file: |
| 6 | ```bash |
| 7 | cat ~/.claude/brain-conflicts.json 2>/dev/null || echo '{"conflicts":[]}' |
| 8 | ``` |
| 9 | |
| 10 | 2. Filter to unresolved conflicts (where `resolved` is not `true`). |
| 11 | |
| 12 | 3. If no unresolved conflicts, tell the user: "No pending conflicts. Brain is fully synced." |
| 13 | |
| 14 | 4. For each unresolved conflict, present: |
| 15 | - The topic |
| 16 | - What Machine A says |
| 17 | - What Machine B says |
| 18 | - The AI's suggestion and confidence level |
| 19 | |
| 20 | Ask the user to choose: |
| 21 | - **Accept AI suggestion**: Apply the suggestion |
| 22 | - **Keep Machine A's version**: Use A's content |
| 23 | - **Keep Machine B's version**: Use B's content |
| 24 | - **Keep both (machine-specific)**: Tag each with its machine name |
| 25 | - **Custom**: Let the user type their own resolution |
| 26 | |
| 27 | 5. After each resolution: |
| 28 | - Mark the conflict as `resolved: true` with the chosen resolution in brain-conflicts.json |
| 29 | - Apply the resolution to the appropriate brain file (CLAUDE.md, memory, etc.) |
| 30 | |
| 31 | 6. After all conflicts are resolved: |
| 32 | ```bash |
| 33 | bash "${CLAUDE_PLUGIN_ROOT}/scripts/push.sh" |
| 34 | ``` |
| 35 | |
| 36 | 7. Show summary: "X conflicts resolved. Brain is now fully synced." |