Claude 4eae92964e docs: simplify CONTRIBUTING-WITH-AI around human review and on-device testing before any PR ...

Cut the doc from 411 to ~75 lines. The performance, coroutine, caching,
logging and KMP rules already live in CLAUDE.md and the skills that agents
load, so they are dropped here. The doc now leads with one rule: no PR is
opened until a human has reviewed the diff and tested the app by hand on a
device, and an agent never opens the PR itself. CONTRIBUTING.md, the PR
template and CLAUDE.md now point to that rule.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01DbNeTtN2RpXHeGwNzPfvbQ

2026-10-06 19:46:23 +00:00

..

actions

fix(ci): make the Homebrew/Winget bump workflows actually fire

2026-07-29 09:37:24 -04:00

ISSUE_TEMPLATE

Update bug_report.md

2025-04-12 15:56:51 +01:00

workflows

docs(release): treat Homebrew as autobumped; cask script verifies instead of submitting

2026-10-06 10:10:16 -04:00

dependabot.yml

feat(release): expand desktop distribution to 8 assets + 2 package managers

2026-04-16 14:54:51 +03:00

PULL_REQUEST_TEMPLATE.md

docs: simplify CONTRIBUTING-WITH-AI around human review and on-device testing before any PR

2026-10-06 19:46:23 +00:00