Use whenever the user wants to analyze or understand what happened in an Inspect AI eval, sample, or set of logs. Routes to the right tool (`inspect_ai.log` for single-log work, `inspect_ai.analysis` for cross-log dataframes, Inspect Scout for transcript pattern detection) and covers the `inspect_ai.analysis` surface for ad-hoc analysis. Triggers on questions like "what happened in this eval/sample", "how did model X compare to Y", "show me outliers across these runs", "did any sample do X", or any follow-up where the user is analyzing logs (not just reading them).
Use for monitoring and diagnosing RUNNING Inspect AI evaluations via the `inspect ctl` command-line control channel. Triggers when: watching a live eval for stalls/errors/retry failures, checking progress of an active eval process, launching an eval you'll monitor in parallel, or finding problematic samples in a running eval.
Use whenever the user needs to read, inspect, or process Inspect AI eval log files (`.eval` or `.json`). Covers the Python API (`read_eval_log`, `read_eval_log_samples`, `read_eval_log_sample_summaries`, `list_eval_logs`), `header_only` mode, the critical anti-pattern of unzipping `.eval` files, and memory-safe patterns for large logs.
Use whenever the user is working with Inspect AI or any of its ecosystem packages (Evals, Flow, Scout, Viz, SWE, Harbor, Sandboxes). Maps concerns to packages and points at each package's docs before calling its API.