Tasks: the start opens the conversation

The last spec artifact. A DAG derived from the design and testing plan; each task names the testing-plan row that proves it. TDD: the test first, red, then green.

Task list

• [x] 1. The channel and the state — channels/slack.py: open(work_item), _open_thread(…, origin); channels/state.py: CONVERSATION_ORIGINS + start; eventlog.py: channel.open_failed, the thread_opened description
  • Depends on: none
  • Requirements: R1.2, R1.3, R1.5, R2.1, R2.2
  • Test: T1 — test_channels.py::test_open_posts_the_root_alone_and_binds_with_origin_start, ::test_open_is_idempotent_for_a_bound_work_item, ::test_open_fails_closed_like_post, ::test_a_failed_open_binds_nothing; T10 — ::test_an_unknown_origin_is_coerced_to_event; test_eventlog.py catalog
• [x] 2. The bus and the opener — channels/bus.py: open_conversation; channels/publishers.py: conversation_opener
  • Depends on: 1
  • Requirements: R1.5, R1.6, R2.2, R3.1
  • Test: T1 — test_bus.py::test_open_conversation_opens_on_every_channel_that_can_and_skips_the_ledger, ::test_a_failing_open_is_a_result_and_an_event_never_an_exception, ::test_the_daemon_opener_reads_the_config_per_call_and_needs_a_channels_section
• [x] 3. The dispatcher seam and the wiring — webhook/dispatcher.py: opener, _open_conversations at the top of _spawn_for; webhook/daemon.py, poller/daemon.py (_build_dispatcher(cli_config_getter)), core/sessions._dispatcher_for
  • Depends on: 2
  • Requirements: R1.1, R1.4, R3.2, R3.3
  • Test: T1 — test_control_integration.py::test_a_start_opens_the_conversation_once_before_the_checkout, ::test_a_refused_start_opens_no_thread, ::test_an_unauthorized_start_opens_no_thread, ::test_a_raising_opener_never_fails_the_spawn, ::test_the_opener_is_handed_the_ref_alone; test_core_sessions.py::test_the_facade_dispatcher_opens_with_the_config_it_was_given
• [x] 4. Scenarios — the four Gherkin scenarios in test_channels_integration.py
  • Depends on: 3
  • Requirements: R1.1–R1.5, R1.7, R2.1
  • Test: T2; T8 — A3 ::test_a_channel_outage_never_fails_the_spawn, A5 test_channels.py::test_a_corrupt_state_file_still_opens_on_start
• [x] 5. Docs, capability docs, decision — docs/capabilities/channels.md, docs/capabilities/interactive-sessions.md, docs/cli/commands/channels.md, docs/cli/commands/sessions.md, docs/cli/state.md, docs/config/cli/channels-options.md, skills/the-loop/reference/collaboration.md, decision-107 + index row
  • Depends on: 3
  • Requirements: the capability-docs gate
  • Test: T12 — make check (markdownlint included)
• [x] 6. Verification — execute testing-plan.md, record evidence/verification.md and evidence/security-review.md
  • Depends on: 4, 5
  • Requirements: all
  • Test: T1, T2, T8, T10, T12, T13

Dependency graph (DAG)

flowchart LR
  T1["1 channel open + origin"] --> T2["2 bus + opener"]
  T2 --> T3["3 dispatcher seam + wiring"]
  T3 --> T4["4 scenarios"]
  T3 --> T5["5 docs + decision"]
  T4 --> T6["6 verification"]
  T5 --> T6

Checkpoints

After task 1 and after task 3: the named tests red → green recorded in evidence/verification.md. After task 4: the integration files green. After task 5: make check. Then the verification node, then the self-review rounds and the security review gate (evidence/security-review.md), then the PR with the reviewer briefing.

Review comments

Appended by the-loop's record-feedback hook when a human gate approves with comments (issue-109). Append-only and attributed: an approval never silently discards a reviewer's suggestions, and the feedback travels with the document it concerns rather than living in a side-channel tracker.