[yasyf]

Raises the spawnllm floor to 0.15.0 (yasyf/spawnllm#11), where the Codex backend's small and medium tiers resolve to gpt-6-luna instead of gpt-5.6-luna. Codex calls already run on service_tier=fast, so every capt-hook LLM judge that routes to Codex (the review and debugging specialties, and any hook that names Codex) now runs gpt-6-luna on the fast tier.

Owner ruling (2026-10-01, 6:3x PM PT): "use a faster LLM for that hook, it should be using gpt-6-luna fast mode by default, that's a change in capt-hook/spawnllm". The monorepo Slack-write hook moves from the general specialty (Claude Haiku) to review in Forge-AI/monorepo#29230.

[socket-security]

Review the following changes in direct dependencies. Learn more about Socket for GitHub.

Diff	Package	Supply Chain Security	Vulnerability	Quality	Maintenance	License
	pypi/​spawnllm@​0.14.0 ⏵ 0.15.0

View full report