The foreman records its judgments in one persistent TASK-LEDGER.md for the task. Its location, schema, ownership, and migration contract are in skills/herdr-foreman/state-schema.md, Task Ledger. The JSON utility ledger continues to record what was dispatched and which recovery decisions were authorized. An applied row there says nothing about whether the work passed.

Start or Resume

At Step 3, establish the stable task identity, original base, and absolute ledger path. Save that path beside the task's authorization context and include it in every foreman handoff. Create the document on a first run, or read the existing document in full before planning more work. Do not create a new ledger per fix.

Read the referenced utility state through its owner commands. Correlate each assignment with its dispatch ID, worker, role, and fresh report path. Never edit state.json by hand to make it agree with a Markdown decision.

On resume, inspect evidence relevant to the next action: the actual report and its content, branch tip/diff, review and test results, release state, and live pane/native evidence for work that may still be running. Stale Herdr working does not undo accepted work; idle or done, however often observed, does not accept work. A conflicting observation requires reconciliation, not a resend.

Record Decisions as They Happen

Append an event after each dispatch outcome, wait outcome, and evidence-based assessment. Record missing reports and unknown send outcomes before continuing another worker. Append blocker, judge, release, cleanup, and handoff decisions when they happen; Step 16 finalizes the log rather than creating it from memory.

Pick each event's decision from its subject's vocabulary in DECISION_MEANINGS at the top of skills/herdr-foreman/foreman/members.py, which also gives each decision's meaning. Task and assignment decisions are separate vocabularies.

Record Herdr's label under observed, with its source. Record worker claims as claims there too. Neither is the foreman's decision. found: true confirms report delivery only. A report containing ## BLOCKED is not successful task completion. A reviewer assignment can be accepted as a completed review while its blocking findings keep the task in_progress.

Bind every acceptance to the specific report content and artifact inspected. Record the full SHA for repository work and link test/review evidence for that SHA. Preserve evidence files for resume. A new push makes older verification insufficient for release: append that invalidation, put the task back in progress, and obtain Step 12's current-tip verification. Keep the old report and decision.

Verify release claims against the live PR, merge, and any required publication checks under rules/ci-safety.md. An accepted release report is not a substitute for those checks. Record remaining obligations explicitly before claiming the task complete. The ledger never bypasses dispatch readiness or permits input into a working or blocked worker.

Blank Document and Event

Fill this as a reasoning document; placeholder values are not evidence. Append one uniquely named event section per decision. Each placeholder names the skills/herdr-foreman/foreman/members.py constant that fixes its format; write the constant's value, never its name.

---
schema_version: <LEDGER_SCHEMA_VERSION>
task: <stable task identifier>
base_revision: <original commit, matching LEDGER_SHA>
dispatch_state: <absolute path of the utility state.json>
---

# Task Ledger

## <unique event id>

- schema_version: <LEDGER_SCHEMA_VERSION>
- id: <same unique event id>
- at: <observation time, timezone-qualified>
- subject: <one of SUBJECTS>
- dispatch_id: <actual dispatch id, or NOT_APPLICABLE per ASSIGNMENT_IDENTITY>
- worker: <worker name, or NOT_APPLICABLE per ASSIGNMENT_IDENTITY>
- role: <role, or NOT_APPLICABLE per ASSIGNMENT_IDENTITY>
- report: <absolute path, or one of REPORT_PLACEHOLDERS>
- observed: <source and its actual observation>
- decision: <one of DECISION_MEANINGS[subject]>
- head_revision: <inspected commit matching LEDGER_SHA, or one of HEAD_PLACEHOLDERS>
- evidence: <paths/content/digests, refs, gate URLs/results, or UNKNOWN>
- assessment: <reason, remaining criteria, next action>