agnostic-ai v0.67.0: Antigravity loads your rules again Agnostic-ai released v0.67.0, which restores Antigravity rule loading by writing a `trigger` frontmatter line on every rule, moving the entry point to `.agents/AGENTS.md`, and emitting Kilo hook specs as `.kilo/plugin/.ts` modules. The release also fixes scoped rules so they reach Antigravity at `/.agents/rules/.md` instead of being skipped, per issues #1113, #1114, and #1105. Users who sync the `antigravity` target need to run `agnostic-ai sync` and commit the entry-point move plus the new `trigger` lines in `.agents/rules/`; the project notes Antigravity 2.17.0 gives rules a 20,000-token budget and loads repository settings from `.gemini/config.json` instead of `.agents/settings.json`, which agnostic-ai does not write. What to do 1. If you sync antigravity , run agnostic-ai sync and commit the move of the entry point to .agents/AGENTS.md , plus the new trigger lines in .agents/rules/ one sync and one commit . Nothing else needs action. What changed - Antigravity loads every rule again: each one starts with trigger frontmatter taken from alwaysApply , globs , and description 1113 . - A scope d rule reaches Antigravity at