2026-09-23 — The product thinker ruled on the owed items of the autonomous run of 2026-09-23 in one interview, one question at a time (07:51Z to 09:32Z, while the run was paused): two release blockers, two intents, the 35 major captures routed to the product thinker, and two older owed rulings. The release waits for two things: version-stamped plugin payloads (iss-2609230733497536, built by a run lane, with no criterion amended and no deferral) and a release press release composed at the cut from the shipped intents' press releases, which is a new capability and gets an intent and its planning interview before any lane builds it. The run may approve the release environment itself once every gate is green; that agenda line is written in the release commit, not here. itd-4's AC3 is built as written, the `--issue-drift` check and the rename of the back-links to `related_issues` / `related_intents` together, and spc-6 closes only when both land (the 2026-07-17 ruling stands). itd-43's third criterion is moot and dropped, so spc-8 closes. The guard contract is extended to read a host-supplied per-call working directory (iss-2609212142557657, option A), after a probe of what the host does when none is supplied. Four majors are ruled to a build (iss-2608231120121681 write-path smoke, iss-2608290810036869 preflight before the push connects, iss-2609090828371674 per-agent staging locks, iss-2609120505141653 the provenance comment) and do not hold the tag. The rulings that set a rule each have their own entry below; every other routed major is deferred past v0.9.0 with its ruling as the `deferral_reason`, and the 168 routed minor and nitpick captures take the run's default deferral. The product thinker also asked that the interview be kept as autonomous-run experiment data, towards making a run transparent to the person it serves; the observations feed the run's end-of-run research note. Rejected, as the frame: the run taking any of these rulings itself.

Frontmatter

.abcd/work/decisions/dcn-2609230000000001-the-product-thinker-ruled-on-the-owed.md
open on GitHub ↗ · commit history ↗

mentions11

• iss-2608231120121681a lint-and-test-green write path can still be broken; nothing requires running the built binary
• iss-2608290810036869git push can report success while pushing nothing, when the pre-push preflight outlasts the SSH idle timeout. The hook runs the full preflight, which takes minutes, and the connection is opened before the hook runs, so the server closes it mid-hook: the push dies with a connection-closed message, the remote ref is unchanged, and the preflight's own passing output scrolls past the failure so the whole thing reads as a clean run. Hit twice in one session. The second time the push was also piped to another command, which returns that command's exit status, so a failed push reported zero. Remedies that worked: set SSH keepalive options for the push, never pipe the push, and confirm with a remote ref listing that the tip actually moved. Worth considering whether the hook should run before the connection is opened, or whether a push wrapper should verify the remote tip afterwards; the silent half is the defect, not the slowness.
• iss-2609090828371674The shared file lock's backoff cap limits staging to roughly ten writers a second, so a burst of simultaneous sub-agent completions loses transcripts. The lock helper backs off exponentially to a hundred-millisecond ceiling, so the rate a contended lock admits is set by that ceiling and not by how short the critical section is. Measured here, sixteen simultaneous stages take 2.12 seconds, consistently across three runs. Extrapolating the same rate, a burst past roughly forty to fifty simultaneous completions begins exceeding the five-second staging lock timeout, and a stage that times out is refused: the transcript it carried is not written anywhere, which is the loss the capture work exists to prevent. Sessions that fan out widely are exactly the sessions whose delegated reasoning is most worth keeping, so the ceiling bites hardest where the value is highest. The condition is pre-existing in the locking helper rather than introduced by sub-agent capture, but nothing reached the lock concurrently before, so it was unreachable in practice until now. The remedies are design-shaped and should be chosen rather than assumed: raise the timeout, lower the backoff ceiling, sharded locks keyed per agent, or a lock-free append with reconciliation at drain.
• iss-2609120505141653`origin: researcher-authored` states that a person invoked the verb. It is
• iss-2609212142557657The guard's hook payload omits a per-call working directory, so a command moved into it slips the command-string match. guardHookInput reads the session cwd and tool_input.command and nothing else; a host whose shell tool takes a per-call working-directory field (opencode's does; a future Claude Code field would) can carry a destructive command whose target is the directory rather than an argument, and the match over argv sees nothing hazardous. Live-demonstrated in lab 2's smoke (probe p5b-live-session-retest under the lab home): the model set the tool's workdir and ran the bare destructive verb. Verified at main on 2026-09-21: the struct is unchanged. Decision owed to the product thinker: extend the guard contract to read a working-directory field where the host supplies one and resolve the command against it, or document the reach in adr-42's mistake-filter scope as outside the boundary. Not a security boundary by adr-42's own statement, which is why the severity is minor; the category is security because the slip is the shape a mistake filter exists to catch.
• iss-2609230733497536itd-67 ac-3 promises that a ship bumps plugin.json.version, updates marketplace.json and that /plugin update abcd pulls the new version; delivered reality: launch ship stamps the version only into a payload staged outside the repository when --payload-dir is passed (internal/surface/cli/ship.go), release.yml uploads binaries and checksums.txt only, and the marketplace source ./ is the git tree whose committed manifests carry no version — so the version-stamped manifests are produced by nothing in the release path and consumed by nothing; the installed plugin's version arrives through the bootstrap binary download alone. Found by the itd-67 fidelity audit (rcp-7af1556ce4f7).
• itd-4Nothing You Notice Gets Lost
• itd-43abcd Speaks One Word for a Specced Block of Work, and That Word Is "Spec"
• itd-2609231013154443Every release arrives with a press release composed from what shipped
• spc-6issue-capture
• spc-8epic-to-spec-terminology