agentleFS
probabl-aiprobabl-ai/skore-skills/skills/review-ml-choices/SKILL.md

review-ml-choices

Skill145 starsChanged 7 days agoBSD-3-ClauseView on GitHub

Show stored project choices and re-enter the skill that can change one. Trigger when the user asks what we decided, what the current settings are, or to change a stored project choice.

  • Reads credentials

1 other project uses a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

7 rules. About 0 are common, 1 is distinctive, and 6 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈1

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 6

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Run python -m skore_skills review choices.

    line 34
  2. Very specific:

    Do not policy set. Do not run the child's commands from memory.

    line 53
  3. Very specific:

    Do not policy set any key.

    line 72
  4. Very specific:

    Do not edit JOURNAL, including a Data understanding skipped row.

    line 73
  5. Very specific:

    Do not load sync-ml-reports when policy.skore_mode is unset.

    line 75
  6. Very specific:

    Do not env add-skore, git commit, or scaffold src/.

    line 77
  7. Distinctive:

    If the owning skill is not installed, show the value and skip the change in one line.

    line 78

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Review ML Choices

Show stored choices. A change loads the skill that already owns that decision. Do not write .skore or the journal from here.

Human-facing prose

Details: setup-workspace references/human_facing_prose.md. The board uses data-science labels: where reports go, executed notebooks, documentation site, git commits, who manages the environment, data analysis, and each filled framing cell (prediction goal, deployment, metric, fold count, and the others below). Do not put skill ids, G-* names, or the wrapper CLI in the question.

Sequence

  1. Run python -m skore_skills review choices. The JSON is the board. Do not rebuild which rows are offered. Do not run frame show or frame clear here.
  2. AskUserQuestion, one pick. Say first, in 2–4 lines, what a change authorizes — re-entering that setup, not a silent flag flip — and echo the current values from the JSON. A file link is an addition, never the context.

    Options are Keep these plus one option per changeable row and one option per framing row. read_only and not_offered are context, not options. Use each row's value. For not_offered, say that row's reason. When framing_reason is set, say it and do not offer framing rows. "Who manages the environment" is whether this project manages it (policy.env.managed). An installed skill is not that value. Do not write a skill id on the board.

  3. Keep these → stop. Do not load a skill.
  4. One changeable row → load that row's skill and stop. The close names the decision being re-opened, not the skill id. Do not policy set. Do not run the child's commands from memory.

    • skore_mode → the user asked to change where reports go.
    • notebooks_site → notebooks and the documentation site.
    • git_autocommit → the user asked to change that choice.
    • env_managed → do not run env add-skore here. If the recorded destination is hub or mlflow, do not use --mode local.
    • data_analysis when action is rerun → do not write a JOURNAL skipped row. The written analysis stays; the change is to run it again.
  5. One framing row → load frame-ml-problem naming that row's id. Do not blank the cell here.
  6. A request to change a read_only row: say that row's reason and stop. Do not load setup-workspace to rename. Do not policy set.

Stop conditions

  • Do not policy set any key.
  • Do not edit JOURNAL, including a Data understanding skipped row.
  • Do not load sync-ml-reports when policy.skore_mode is unset.
  • Do not env add-skore, git commit, or scaffold src/.
  • If the owning skill is not installed, show the value and skip the change in one line. Do not invent that skill's steps.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.