coding.sh score

coding.sh score <action>

  Add and list typed scores on sessions, events, cases and runs

Commands
  add           Score a session, event, case or run (prints the score id)
  list          List scores, newest first
  config        Fix a score name's type and allowed values (prints the config id)
  configs       List score configs

Notes
  Labels are human scores: a pass or fail label is a boolean score, and judge calibrate reads human boolean scores as labels.
  A value outside its config's range or categories is refused (exit 2). The viewer and serve take the same scores at /api/scores.
  A stored score keeps its type: a value that reads as another type is refused (exit 2) unless --type is given or the name has a config.
  An event must be in the store to be scored. Its session is its own: --session with another session is refused (exit 2).

coding.sh score add <TARGET> <NAME> <VALUE> [options]

  Score a session, event, case or run (prints the score id)

Arguments
  target        A session id (a prefix or latest works), or event:<id>, case:<id>, run:<id>
  name          Score name, e.g. outcome or helpfulness
  value         A number, true or false, or a category

Options
  --type <TYPE>              Data type (default: the name's config, else what the value looks like)
  --comment <TEXT>           Why this score
  --source <SOURCE>          Who gave it (default: human)
  --config <ID>              Check the value against this config (default: the config with the score's name)
  --session <ID>             The session a case or run belongs to (an event's own session is looked up)
  --judge-version <VERSION>  Judge version that produced the score
  --id <ID>                  Score id; adding the same id again replaces it (default: one per target, name, source and person)
  --fresh                    Wait for every pending capture to be ingested (default: answer after 1.5 s)

Examples
  coding.sh score add latest outcome true
  coding.sh score add 4f2a helpfulness 4 --comment 'fixed it first try'
  coding.sh score add event:evt_123 correct false --source code

coding.sh score list [options]

  List scores, newest first

Options
      --target <TARGET>  Only scores on this target (as in score add)
      --name <NAME>      Only this score name
      --source <SOURCE>  Only this source
      --session <ID>     Only scores on this session or its events
      --since <WHEN>     Only at or after this time: 7d, 12h, 30m, today, yesterday or a date. A date without a zone is local time: 2026-11-09 is local midnight
  -n, --limit <N>        Maximum rows (default: 50)
      --fresh            Wait for every pending capture to be ingested (default: answer after 1.5 s)

Examples
  coding.sh score list --name helpfulness
  coding.sh score list --session latest --json

coding.sh score config <NAME> [options]

  Fix a score name's type and allowed values (prints the config id)

Arguments
  name          Score name

Options
  --type <TYPE>           Data type (required)
  --min <N>               numeric: smallest allowed value
  --max <N>               numeric: largest allowed value
  --categories <A,B,...>  categorical: the allowed values (comma-separated or repeated)
  --description <TEXT>    What the score means
  --id <ID>               Config id (default: the name)

Examples
  coding.sh score config helpfulness --type numeric --min 0 --max 5
  coding.sh score config tone --type categorical --categories positive,neutral,negative

coding.sh score configs

  List score configs