/just-say-so:summary

What it does

Reads the two state files and prints facts with their counts. The window and its sessions; volume; rejections per 100 turns, overall, per project, and per week once the window spans a month; rejections in each distance bucket from the last rules injection; the rules with an error on final replies; the rules on rejected replies; rewrites; the reasons you typed; the tags, each with what covers it; what each recurring tag points at; allows; sessions per setting. It ends with observations and what the data does not establish. It recommends nothing; advise does.

Synopsis

/just-say-so:summary [--since 30d] [--project <path>] [--json]

Arguments

Argument	Meaning
--since	An ISO date, or a span such as 30d or 6w. Default: all time.
--project <path>	Narrow to one project path.
--json	Print the object the text is rendered from.

What it reads and writes

Reads events.jsonl and rejections.jsonl in the state directory, and the rules file of each project in the report, to classify tags. Writes nothing.

What it prints

The report. It prints rule names, shipped terms, counts, project paths, and your typed reasons. It never prints a prompt or a reply. A rate prints only when its denominator is in the window; otherwise the line says it is unavailable. The definitions behind every number are in Records and the summary.

Example

Tags (what covers each: a judgment bullet, a Vale-backed bullet, or nothing yet)
  hedging: 3 · no bullet covers it yet
  end-with-fact: 3 · covered by a bullet backed by ai-tells.MicDrop, ai-tells.PointVerdict; those rules fired on 0 of 3 linted rejected replies

What the tags point at (per project; a tag needs 3 rejections; ship-time means the reply event, restate-time means /restate linting the old reply)
  /Users/you/proj · #hedging: 3 · uncovered: no bullet has this id. Points at a judgment bullet, or a Vale rule.