report

1# Report
2 
3Report the current evo workspace from recorded state only. A report request is
4read-only, even if the user phrases it casually as "what happened?", "what got
5better?", "what should I pay attention to?", or "I just woke up".
6 
7Do not spend compute while reporting:
8 
9- Do not run `evo run`, `evo gate check`, benchmark commands, or project eval
10  scripts.
11- Do not run `python bench.py`, `python slurm_eval.py`, `sbatch`, `srun`,
12  `squeue`, `sacct`, or `scancel` to verify a result.
13- Do not create launcher, monitor, parsing, or analysis scripts.
14- Do not edit files.
15 
16Use stored evo state instead: `evo report`, `evo status`, `evo tree`,
17`evo frontier`, `evo show <id>`, `evo diff <id>`, and immutable artifacts under
18`.evo/run_*/experiments/<exp>/attempts/<NNN>/`.
19 
20For chart requests, render the dashboard's scatter plot as a colored terminal
21block, one chart per run, sized to the current terminal.
22 
23## What it shows
24 
25Mirrors the web dashboard's score scatter (left rail of `evo dashboard`):
26 
27- X = experiment creation order, Y = score
28- Dot color by status: green = committed valid result, red = failed, purple = active, grey = pending / evaluated / discarded / pruned
29- ★ marks the current best valid committed-result experiment. `pruned` with `prune_kind=exhausted` can still be best; `prune_kind=invalid` and its descendants cannot.
30- Yellow ring on dots that sit on the best-path spine (root → best)
31…