| 1 | # My Tool — Practical Examples |
| 2 | |
| 3 | Concrete scenarios showing when and how to use the my tool effectively. |
| 4 | |
| 5 | ## Diagnosis |
| 6 | |
| 7 | ### "Why can't you search the web?" |
| 8 | ``` |
| 9 | → my(action="check", key="web_config.enable") |
| 10 | → False |
| 11 | → "Web search is disabled. Add web.enable: true to your config to enable it." |
| 12 | ``` |
| 13 | |
| 14 | ### "Why did you stop?" |
| 15 | ``` |
| 16 | → my(action="check", key="max_iterations") |
| 17 | → 40 |
| 18 | → my(action="check", key="_last_usage") |
| 19 | → {"prompt_tokens": 62000, "completion_tokens": 3000} |
| 20 | → "I hit the iteration limit (40). The task was complex. I can ask the user if they want to increase it." |
| 21 | ``` |
| 22 | |
| 23 | ### "What model are you running?" |
| 24 | ``` |
| 25 | → my(action="check", key="model") |
| 26 | → 'anthropic/claude-sonnet-4-20250514' |
| 27 | ``` |
| 28 | |
| 29 | ## Adaptive Behavior |
| 30 | |
| 31 | ### Large codebase analysis |
| 32 | ``` |
| 33 | → my(action="check") |
| 34 | → context_window_tokens: 65536 |
| 35 | → my(action="set", key="context_window_tokens", value=131072) |
| 36 | → "Set context_window_tokens = 131072 (was 65536)" |
| 37 | → "I've expanded my context window to handle this large codebase." |
| 38 | ``` |
| 39 | |
| 40 | ### Switching to a faster model for repetitive tasks |
| 41 | ``` |
| 42 | → my(action="set", key="model", value="anthropic/claude-haiku-4-5-20251001") |
| 43 | → "Set model = 'anthropic/claude-haiku-4-5-20251001' (was 'anthropic/claude-sonnet-4-20250514')" |
| 44 | → "Switched to a faster model for these batch tasks." |
| 45 | ``` |
| 46 | |
| 47 | ## Cross-Turn Memory |
| 48 | |
| 49 | ### Remembering user preferences |
| 50 | ``` |
| 51 | # Turn 1: user says "keep it brief" |
| 52 | → my(action="set", key="user_style", value="concise") |
| 53 | → "Set scratchpad.user_style = 'concise'" |
| 54 | |
| 55 | # Turn 3: new topic |
| 56 | → my(action="check", key="user_style") |
| 57 | → 'concise' |
| 58 | (adjusts response style accordingly) |
| 59 | ``` |
| 60 | |
| 61 | ### Tracking project context |
| 62 | ``` |
| 63 | → my(action="set", key="active_branch", value="feat/auth") |
| 64 | → my(action="set", key="test_framework", value="pytest") |
| 65 | → my(action="set", key="has_docker", value=true) |
| 66 | ``` |
| 67 | |
| 68 | ## Budget Awareness |
| 69 | |
| 70 | ### Token-conscious behavior |
| 71 | ``` |
| 72 | → my(action="check", key="_last_usage") |
| 73 | → {"prompt_tokens": 58000, "completion_tokens": 12000} |
| 74 | → "I've consumed ~70k tokens. I'll keep my remaining responses focused." |
| 75 | ``` |
| 76 |