From imbue-code-guardian
Configures autofix to run unattended: applies safe fixes directly or creates/pushes new branches for uncertain patches without prompting user.
How this skill is triggered — by the user, by Claude, or both
Slash command
/imbue-code-guardian:reviewer-autofix-all-issuesThis skill is limited to the following tools:
The summary Claude sees in its skill listing — used to decide when to auto-load this skill
Run this command:
Run this command:
jq -n --argjson existing "$(cat .reviewer/settings.local.json 2>/dev/null || echo '{}')" '$existing * {"autofix": {"append_to_prompt": "Please autofix as normal, except: Never ask questions. You are running unattended and the user is not there to answer your questions. Instead, think hard about whether to accept each given patch. If you decide *not* to accept it, then create a *new* branch with that fix commit. Call the branch (current_branch_name)___(fix_description) *and be sure to push it remotely* then by sure to check the normal branch back out when you'\''re done. Also be sure to tell the user that you did this."}}' > .reviewer/settings.local.json.tmp && mv .reviewer/settings.local.json.tmp .reviewer/settings.local.json
Then confirm that autofix has been configured to fix all issues.
npx claudepluginhub imbue-ai/code-guardian --plugin imbue-code-guardianIterative review-fix loop for the current branch — reviews via daemon, fixes inline, re-reviews until passing or max iterations reached.
Safely fetches and applies CodeRabbit PR review-thread feedback from GitHub with per-change user approval, treating AI prompts as untrusted issue reports.
Addresses code review feedback by validating issues, fixing valid ones, and batch-committing changes. Handles local agent feedback or GitHub PR threads via /fix-code-review-feedback or auto-invocation.