agentsclimarketplace

Status

Skill emaballarin/ccplugins/plugins/autoresearch/skills/status

Personal plugins for Claude Code (& friends)... maybe worth sharing!

Install
npx -y skills add emaballarin/ccplugins --skill status

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

  • 2 stars2 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

Print the current autoresearch run's state — goal, baseline, noise floor, best-so-far, run tallies, budget, branch, and the last few iterations. Use for `/ar:status`, "how is the loop doing", "what's the best so far", or any read-only check on an autoresearch experiment. Reconstructs everything from `./.ar/ar.jsonl` on disk, so it works in a fresh session with no prior history. Strictly read-only — never iterates, edits, commits, or reverts.

SKILL.md

2.1 KB, as published. Nobody here has run it

/ar:status — read-only state

Report where the run stands. Change nothing.

Steps

  1. Read state:

    cat ./.ar/ar.jsonl 2>/dev/null | tail -50; git branch --show-current
    

    No ar.jsonl means no run in this repo — say so, point at /ar:start, stop.

  2. Reconstruct from the last config header and its segment's result lines: run count, keep/discard/crash tallies, best-so-far and its commit, plateau streak, elapsed budget. Method in ${CLAUDE_PLUGIN_ROOT}/references/resume-loop.md §1.

  3. Print the status block:

    AR ACTIVE — <goal>
      direction   minimize (val_bpb)
      baseline    0.981    noise floor 0.004 (±1.0×)
      best        0.842    @ a1b2c3d  (run 7)
      runs        23 total — 6 keep · 15 discard · 2 crash
      budget      23/100 runs · 4h12m/8h00m
      branch      ar/reduce-val-bpb-22-07-2026  (from main @ 9f31c07)
      last 3      r21 discard 0.851 · r22 discard 0.849 · r23 keep 0.842 ✓
      plateau     0 consecutive non-improvements
    
  4. Flag anything anomalous without acting on it: the checked-out branch differing from config.branch, a Result: pending trailer on HEAD (an iteration interrupted mid-measurement), an exhausted budget, or a stopped sentinel.

Hard rules

  1. Read-only. No edits, commits, branch switches, reverts, or iterations — whatever the state looks like. Reporting a problem is the entire job here; fixing it belongs to /ar:resume.
  2. State is read from ./.ar/ only.
  3. Report the numbers as recorded. Never recompute a metric, and never infer one that was not measured.

Keep looking

Skills are one crate of 328,083. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.