○guaranteed
Sessions and handoverintermediateagentreviewerwrites state
A worker opened a question, ran out of session and handed the work over. The next worker starts a fresh task. If the question belonged to the old task, the new task would be acceptable with the question still unanswered, and the handover would have laundered a blocker into a footnote.
handover --close: the continuation record, and the task marked handed over
start <task> --scope <paths>: the follow-up; the prior handover is named
check: the unresolved question is a failing line on this branch
question list: the entry, numbered, with the task that opened it
finish --outcome completed: refused by the same entry
setup: finish-blocker-open
given:
- 'an active task with one unresolved question open against it'
steps:
- id: hand-over
run: ['handover', '--close']
stdin: handover-body.md
note: 'the record the next session resumes from; --close marks the task handed over so a new one may start'
expect:
exit: 0
stdout_contains: ['\.ai/local/state/handovers/']
- id: follow-up
run: ['start', 'the follow-up', '--scope', 'lib']
note: 'the next task; the prior handover for this worktree and branch is named on the way in'
expect:
exit: 0
stdout_contains: ['^INFO handover', 'same_worktree_same_branch', 'next: majordomus context']
- id: still-blocked
run: ['check']
note: 'the question opened under the previous task still blocks this branch'
expect:
exit: 10
stdout_contains: ['^FAIL blockers', 'unresolved question', 'tabs']
- id: still-listed
run: ['question', 'list']
note: 'the entry, numbered, with the task it was opened under'
expect:
exit: 0
stdout_contains: ['^1 \[unresolved\]', 'tabs']
- id: refused
run: ['finish', '--outcome', 'completed', '--verify-command', 'test -d .ai']
note: 'completion of the follow-up is refused by the same entry'
expect:
exit: 10
stdout_contains: ['^FAIL blockers']
then:
- 'the handover carries the branch and head it was written at'
- 'a question is a fact of the branch, not of the task that happened to open it'
- 'nothing was resolved by handing over; the next worker sees the same blocker'
A blocking question keeps blocking after the work is handed to a new task. The record of who asked and under which task survives, and only an answer clears it.
The scenario
Executed against the real tool, in a repository prepared by finish-blocker-open, every time this site is generated. What follows is not a description of what should happen; it is what happened.
given
- ·an active task with one unresolved question open against it
when
- 1
majordomus handover --close - 2
majordomus start the follow-up --scope lib - 3
majordomus check - 4
majordomus question list - 5
majordomus finish --outcome completed --verify-command test -d .ai
then
- ·the handover carries the branch and head it was written at
- ·a question is a fact of the branch, not of the task that happened to open it
- ·nothing was resolved by handing over; the next worker sees the same blocker
What the tool did
Evidence of the run this page was generated from: pass, 5 step(s). Output is normalised: the repository's path, timestamps, ids and durations are placeholders; nothing else is edited.
Tested by test/cases/94_use_cases.sh, which runs every scenario and refuses a step whose recorded output is not what it produced.
-
the record the next session resumes from; --close marks the task handed over so a new one may start
.ai/local/state/handovers/<time>--main--<head>--<hash>.md
- exit 0
- stdout_contains /\.ai/local/state/handovers//
-
the next task; the prior handover for this worktree and branch is named on the way in
started t-<id> profile=implementation scope=lib
INFO handover .ai/local/state/handovers/<time>--main--<head>--<hash>.md — prior record, same_worktree_same_branch, exact, <age> ago [reproduce: majordomus handover --resolve]
next: majordomus context; checkpoint every 15m; majordomus check before claiming anything
- exit 0
- stdout_contains /^INFO handover/
- stdout_contains /same_worktree_same_branch/
- stdout_contains /next: majordomus context/
-
the question opened under the previous task still blocks this branch
OK knowledge records — 0 record(s) valid against knowledge/v1; every reference resolves [reproduce: majordomus knowledge check]
· checkpoint t-<id> — <timed against the checkpoint interval>
OK records ledger.jsonl — every line is a well-formed event
FAIL blockers t-<id> — 1 unresolved question(s) on this branch; first: t-<id> — does the parser need to accept tabs? [reproduce: majordomus question list]
OK records decisions.md — every entry is attributable
OK scope t-<id> — 1 touched file(s), all within scope
OK state t-<id> — exact (head <head>)
INFO mesh t-<id> — the repository declares no enabled mesh
INFO use-case 30 advisory gap(s) — not covered by an active use case, and the policy does not require it: command:init command:update command:doctor command:watch command:start command:context command:checkpoint command:handover command:check command:finish command:history command:decision command:question command:prompt command:evidence command:skills command:capture command:search command:doctrine command:plan command:version command:session command:knowledge command:adr command:usecase command:rules command:migrate command:recover command:bench command:archive [reproduce: majordomus usecase coverage]
OK use-case 0 of 30 target(s) covered — every public command the policy requires is named and run by an active use case (policy use_cases.coverage)
INFO obligation t-<id> — the task declares no obligations
INFO gate t-<id> — this repository declares no CI model at .ai/repo/ci/gates.yaml; nothing can be planned
check: 12 finding(s), 1 failing
- exit 10
- stdout_contains /^FAIL blockers/
- stdout_contains /unresolved question/
- stdout_contains /tabs/
-
the entry, numbered, with the task it was opened under
1 [unresolved] t-<id> — does the parser need to accept tabs? (<date>)
- exit 0
- stdout_contains /^1 \[unresolved\]/
- stdout_contains /tabs/
-
completion of the follow-up is refused by the same entry
FAIL blockers t-<id> — 1 unresolved question(s) on this branch; first: t-<id> — does the parser need to accept tabs? [reproduce: majordomus question list]
FAIL note t-<id> — no --note file and no handover for this task [reproduce: majordomus handover < note.md]
INFO regression t-<id> — profile implementation adds no requirement beyond the shared contract
OK scope t-<id> — 1 touched file(s), all within scope
OK state t-<id> — exact (head <head>)
INFO continuity t-<id> — applies to partial and blocked
INFO mesh t-<id> — the repository declares no enabled mesh
INFO use-case 30 advisory gap(s) — not covered by an active use case, and the policy does not require it: command:init command:update command:doctor command:watch command:start command:context command:checkpoint command:handover command:check command:finish command:history command:decision command:question command:prompt command:evidence command:skills command:capture command:search command:doctrine command:plan command:version command:session command:knowledge command:adr command:usecase command:rules command:migrate command:recover command:bench command:archive [reproduce: majordomus usecase coverage]
OK use-case 0 of 30 target(s) covered — every public command the policy requires is named and run by an active use case (policy use_cases.coverage)
OK verification t-<id> — test -d .ai — exit 0, <s>s, tree <tree>
INFO obligation t-<id> — the task declares no obligations
INFO gate t-<id> — this repository declares no CI model at .ai/repo/ci/gates.yaml; nothing can be planned
INFO gate t-<id> — not in verification.finish_requires
finish: refused, 2 unmet
blocking doctrines:
- majordomus.blocker-resolution
- majordomus.note-integrity
- exit 10
- stdout_contains /^FAIL blockers/
Why the tool knows this
The chain from this page to the code: the use case names commands, rules and claims; each rule is dispatched from the registry and each claim names the test that proves it; the scenario above executed the commands. Nothing in this chain is prose about intent.
The rules that make it hold
Guarantees exercised
Related use cases
Computed from what they share: claims, rules, commands, category, applications. Nobody maintains this list.