Session Handoff [auto-1987]: Guard ruleset_id()'s ids= Extraction Against Malformed Array Elements (#1987) #2035

Next steps, in priority order

1. Fix ruleset_id()'s ids= Extraction Still Misreads a Malformed Ruleset Array Element #1987 in repo-config/configure.sh's ruleset_id(). Guard the ids= extraction the same way ruleset_id()'s per_page Guard Fails Open on a Non-Array gh api Response #1971 (PR Validate the Rulesets Response Before ruleset_id() Reads It (#1971) #1988) already guards the out= read and the array-type check: when the jq extraction fails (an array of non-objects), print a context line and return non-zero rather than falling through to the empty-means-not-found return. Also treat a resolved id of literal null (a matching object with no id field) as unreadable and return non-zero, so no caller reaches repos/<repo>/rulesets/null. Done looks like: both shapes the issue names make ruleset_id fail loud, a well-formed list still resolves as before, and a test in the configure test harness covers each shape, proven by reverting the fix and watching it fail.
2. Drive the pull request to develop per drive-pr, with local-strict-review before every push, the body carrying Closes on promotion: #1987.
3. Close this lane out: comment what merged and what was filed, then close this handoff with no successor.

External blockers

None known at pick time.

Internal dependencies

Step 2 follows step 1, and step 3 follows the develop merge.

State

• No branch, worktree, or pull request exists for this lane yet. The worker creates feature/auto-1987 from develop.
• At pick time develop and main held the same commit, so nothing was waiting on promotion.
• ruleset_id()'s per_page Guard Fails Open on a Non-Array gh api Response #1971 is merged (PR Validate the Rulesets Response Before ruleset_id() Reads It (#1971) #1988), and the current ruleset_id() on develop still assigns ids= unguarded, so the defect is live.

The parked decision queue

Zero parked decisions on this lane.

What the last round did

Nothing on this lane yet. The unattended picker created this handoff from the open backlog, choosing #1987 as a correctness defect in a script that runs, small and self-contained, with a concrete fix named in the issue and no open work naming it.

What not to repeat

Nothing yet.

New learnings

None yet.