v5.166.5

Fixed

• The close-on-merge ledger sweep no longer re-reads every merged-PR claim's issue from GitHub on every governor pass (#11778). On a hive with ~1.1k merged claims that was ~3.4k charged GET /repos/{o}/{r}/issues/{n} an hour attributed to hive:close_on_merge, enough to push the core budget to critical twice in 90 minutes, stretch the eval ticker, and send stop-polling nudges to agents. A claim now records close_on_merge_settled_at once the sweep reaches a final verdict and is skipped with no API call until issues.close_on_merge_recheck (default 24h) elapses; transient read/write failures still retry. The full-ledger walk runs on its own issues.close_on_merge_interval (default 10m) instead of on every pass, passes in between only check claims whose PR merged since the last walk (so a fresh merge still closes its issue promptly), and both the ledger sweep and the backfill pause while the API budget mode is not normal, with one log line per pause/resume. Env overrides: HIVE_CLOSE_ON_MERGE_INTERVAL, HIVE_CLOSE_ON_MERGE_RECHECK.

---

SBOMs: SPDX JSON, generated by Syft against the linux/amd64 image digest, attached below for hive, hive-contributor, and hive-hub. They cover linux/amd64 only — see releases.md for why linux/arm64 is not scanned separately.

NOTICE: third-party license attribution for the Go module dependencies compiled into these images, attached below. Does not cover base-image OS packages or the Node.js/tmux layers built from source — see releases.md for the exact boundary.