Execution Log: the-loop opens a pull request and tells nobody it opened it

Append-only log of progress for the user's visibility.

Phase transitions

Phase	Entered	Reviewed/approved by	Notes
requirements-definition	2026-08-20		bugfix.md derived from the ticket; the owner had already chosen fix 1 on the ticket
design	2026-08-20		design.md
test-planning	2026-08-20		testing-plan.md
tasks-breakdown	2026-08-20		tasks.md
implementation	2026-08-20		tasks 1–7
verification	2026-08-20		every activity in testing-plan.md ran; results and evidence recorded there
needs-review	2026-08-20		reviewer briefing posted on PR #276
complete

Pull requests

PR	Scope / tasks	Status
#276	the whole work item — spec chain, the link-pr operation on its surfaces, the workflow rule, documentation and tests	open

Progress entries

2026-08-20 — implemented, verified, ready for review

• Phase: implementation → verification → needs-review
• Did: red run first (19 tests, 18 failing on the missing operation, 1 control holding the bug in place); then core.sessions.link_pull_request and its surfaces (CLI sessions link-pr, POST /api/v1/sessions/link-pr, the link_pull_request MCP tool, sessions.link_pr on the SDK), the authored OpenAPI contract, the workflow rule in the four places the labelling rule already lives, the capability/CLI/state/SDK docs and decision-098.
• Checkpoint/tests: make test — 2495 passed, 1 skipped (2476 on main at 50c2a27, +19); 2501 passed, 1 skipped after rebasing onto main at 71e7dff. make lint, make format-check, pyright cli, validate_config.py — all clean first run. Evidence under evidence/.
• Next: human review of PR #276.

2026-08-20 — rebased onto main at the author's request

• Phase: needs-review
• Did: rebased onto main at 71e7dff. One conflict, in docs/capabilities/webhook-triggers.md's history table: issue-273 added its row at the top while this branch added issue-274's. Both kept, newest first. Recorded in bugfix.md that the companion bug named in § Out of scope has now landed as issue-273 (#275), and that it does not close this one.
• Checkpoint/tests: make test — 2501 passed, 1 skipped. make lint, make format-check, pyright cli, validate_config.py — clean.
• Next: human review of PR #276.

2026-08-20 — spec chain locked

• Phase: requirements-definition → tasks-breakdown
• Did: read the ticket and the owner's answer (fix 1); traced the four linkage sources in webhook/router.py and sessions/registry.py and confirmed the-loop's spec PR satisfies none of them; wrote bugfix.md, design.md, testing-plan.md, tasks.md.
• Checkpoint/tests: none yet — no production code changed.
• Next: task 1, the red run.

Verification results

This work item has a testing-plan.md; results are recorded there.

Design critic review

Not selected for this work item.

Review cycles

Cycle	Type (self/critic/security)	Reviewer	Outcome	Link
1	self	the-loop	new findings — the SDK namespace was missing link_pr, and the testing plan named a parity test (test_harness_usage.py) that is not one	this PR
2	self	the-loop	zero (converged)	this PR
3	critic	—	unavailable: reviews.critics is empty in this repository's harness config, so no critic harness is configured to run	.the-loop/harness-config.yaml
4	security	the-loop (checklist)	pass, no findings	evidence/security-review.md

Security review (gate)

• Mechanism: the-loop checklist (security.review.mechanism: auto, the skill being unavailable in this session)
• Outcome: pass, no findings — evidence/security-review.md
• Human sign-off: n/a — effective risk tier 3, below security.review.humanSignOffMinTier (4)

Final validation evidence

The acceptance criteria, mapped onto what proved them (raw record: testing-plan.md § Verification results):

Criterion	Proved by
R1.1 — one implementation, every surface	test_sessions_command_link_pr_* (CLI), test_api_contract_parity + test_api_routers_integration (HTTP), test_mcp_integration (MCP), test_p2_every_namespace_method_reaches_core (SDK)
R1.2 — the PR is recorded and session.pr_linked emitted	test_link_pull_request_records_the_pr_and_emits
R1.3 — idempotent	test_link_pull_request_is_idempotent
R1.4 — no record, no write, exit 1	test_link_pull_request_without_a_session_record_writes_nothing
R1.5 — a work item does not deliver itself	test_link_pull_request_refuses_a_work_item_delivering_itself
R1.6 — bare number, #N, and a cross-repository full ref	test_link_pull_request_resolves_a_number_in_the_work_items_repository (4 cases), test_link_pull_request_accepts_a_pull_request_in_another_repository
R1.7 — malformed input refused, nothing written	test_link_pull_request_refuses_malformed_input (6 cases)
R2.1–R2.3 — the workflow rule, in four places, best-effort	the diff to skills/the-loop/reference/automation.md, commands/work-on.md, commands/execute-tasks.md, skills/the-loop/templates/execution-log.md
R2.4 — the capability doc documents a binding the-loop writes	the diff to docs/capabilities/webhook-triggers.md and its history row
R3.1 — a regression test per layer	evidence/red.md (18 red) → evidence/unit-and-integration.md (all green)
R3.2 — the reproduction end to end, Gherkin	test_a_review_comment_on_the_loops_own_spec_pr_is_lost_without_the_binding and …_reaches_the_session_once_recorded

Capability docs

Capability doc	What changed	History row
docs/capabilities/webhook-triggers.md	the durable PR→work-item binding documented as a source the-loop writes at PR-open time, not only one the dispatcher infers; session.pr_linked names its second writer	issue-274
docs/capabilities/cli.md	sessions link-pr added to the CLI surface, with what it refuses	issue-274

Documentation

Document	What changed
docs/cli/commands/sessions.md	the link-pr action, its flags and when to run it
docs/cli/service.md	link_pull_request in the routed-operations table
docs/api-specs/openapi/the-loop.v1.yaml	POST /api/v1/sessions/link-pr and SessionLinkPrBody
skills/the-loop/reference/automation.md	the PR-open step: record every PR you open
commands/work-on.md, commands/execute-tasks.md	the same rule, beside the labelling rule
skills/the-loop/templates/execution-log.md	the Pull requests note carries the rule
docs/cli/state.md	pullRequests has a second writer: the session that opened the PR
docs/sdk/reference.md	sessions.link_pr
docs/decisions/decision-098.md, docs/decisions/decisions.md	the decision record and its index row