anchors review

Record a review of a target, or list what is to review.

A gate that declares `review:` marks its targets TO REVIEW, apart from how it
measures: a target is to review until a review is recorded at its current revision, and a
change to it makes the review due again. Reviews inform; they never block.

  anchors review --pending [--gate <g>]                 what is to review, and the question
  anchors review <target> --gate <g> --by <who>         a review that found nothing
  anchors review <target> --gate <g> --by <who> --findings "<report>"

--by says who reviewed, as you name them (human:ana, agent:<vendor>/<model>); the record
keeps it. --findings is the whole report — each finding with what, WHERE (file:line) and
why — and it opens an issue, the way a failing judgment does: each finding is then a
failing test and a fix, or dismissed with its reason in the issue.

Usage

anchors review [target] [flags]

Flags

Flag	Default	What it does
--by		who reviewed, as you name them (human:ana, agent:/)
--findings		the findings, as the whole report; each one opens the issue’s work
--gate		the gate whose review this is (a gate that declares review:)
--map		path to the map
--pending		list the targets to review
--record-issues		in manual mode, also write the issue of the findings
--root	.	project root

Judgment and review — family

Commands of the same family — what tells each apart:

Command	What it does
anchors judge	Record an AI’s verdict for a judgment gate
anchors review	Record a review of a target, or list what is to review
anchors suggest	List, apply and decide the proposed fixes

---

---

---

---

---

A gate that declares review: marks its targets to review, apart from how it measures. A target stays to review until a review is recorded at its current revision; a change to it makes the review due again. Reviews inform and never block.

# What is to review, grouped by gate, with the question
anchors review --pending [--gate <gate>]

# A review that found nothing
anchors review <target> --gate <gate> --by human:ana

# A review with findings: the whole report, each finding with what, where and why
anchors review <target> --gate <gate> --by agent:<vendor>/<model> --findings "<report>"

Flag	What it does
--pending	lists the targets to review
--gate	the gate whose review this is — a gate that declares review:
--by	who reviewed, as you name them; the record keeps it and does not grade it
--findings	the findings report; it opens the target’s issue for the gate
--record-issues	in manual mode, also write the issue of the findings

The product of a review is findings, not a stamp: each one becomes a failing test and a fix, or is dismissed with its reason in the issue. There is no waived — a review that did not happen is still due. See judgment and review.