Reply shapes reach the reply, a Stop hook checks completion reports, and hook pipes keep their output #156
No Assignees
Notifications
Due Date
No due date set.
Dependencies
No dependencies set.
Reference: bvandeusen/FabledScribe#156
Reference in New Issue
Block a user
Six commits from dev. They finish milestone 410 "One owner per piece of guidance" (step 8) and build steps 4–5 of milestone 409 "Response shapes — replies say where the work stands", with two hook fixes found along the way.
Guidance
Response shapes (milestone 409)
update_taskreturns matching preferences asreply_preferences. It uses a kind-filtered preference search keyed on each preference's own trigger and logs underreport_preferencein retrieval_logs.scribe_report_check.shchecks the final reply says where the work sits, what needs the operator, and what comes next.GET /api/plugin/report-check.plugin, actionreport_check) for the milestone's step 6 to read.Fixes
tail | grep -qunder pipefail failed exactly when it matched. Test transcripts also weren't compact JSON.$(… | head -N) || v=""discarded head's output when the producer was still writing, so the by-name duplicate arm went silent for heavily duplicated names. Fixed at five sites, and check_plugin now carries a known-bad pattern for the shape.Verification
0fab082(run #519): lint, typecheck, plugin hooks, Python tests, integration, image build.2026.09.14.2254.🤖 Generated with Claude Code
Everything else Scribe gives an agent arrives before the reply is written. A Stop hook is the one moment the finished reply exists, so it is the last chance to fix a report the operator can't read, and the only place adherence to the shape can be measured. - plugin/hooks/scribe_report_check.sh (Stop): deterministic, no model call. 1. Did this turn close a task? That means an update_task/create_task call with status "done" since the turn's prompt, whose tool_result is not an error. Otherwise it stays silent, which covers most turns (one grep). 2. Does the reply that ends the turn say where the work sits (a record by id and title, or step N of M), what needs the operator, and what comes next? Matched on those words, not on exact headings. 3. If sections are missing, it blocks once. With stop_hook_active set, a rewrite is recorded (passed_after_rewrite / missing_after_rewrite) and never blocked again. A block loop started by another plugin (no marker from this hook) is left alone. - Measured: every checked reply is reported to GET /api/plugin/report-check (passed / blocked / after rewrite). Turns that close nothing are not reported; they would cost a request per turn and add nothing to the rate. Outcomes go to app_logs as category "plugin", action "report_check". - It blocks only when the block was recorded, and only in the server's words. The endpoint returns the block reason, so the hook carries timing and transport only (PACKAGING.md), and an unconfigured or unreachable instance never stops a session. - The transcript format is read from real transcripts and marked in the hook as observed rather than documented. The Stop contract (transcript_path, stop_hook_active, decision/reason, no matcher, SubagentStop separate) was checked against the Claude Code hooks docs. A prompt-type hook was not needed: the deterministic check passed a real completion report from this session and blocked a stripped one. - A pipefail trap was caught while exercising the hook: `tail | grep -q` reports failure exactly when grep matches, because tail dies of SIGPIPE. The prefilter reads through process substitution; the section checks use here-strings. - Tests: an end-to-end hook suite over synthetic transcripts and the shared HTTP sink (silence, pass, server-worded block, rewrite recorded, foreign loop, errored write, earlier turn, unwritten reply, no recorded check, bare id), and service tests for the reason wording and the outcome record. Smoke event added to check_plugin; README and PACKAGING list the hook and endpoint. Plugin version minted. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>