1password
Install and use 1Password CLI (`op`) to sign in, inspect vault items, read secrets safely, and inject secrets into commands.
Review diffs and change sets for bugs, regressions, risks, and missing tests.
$ npx -y skills add HybridAIOne/hybridclaw --skill code-review --agent claude-codeHow it fires
How this skill gets triggered: by you, by Claude, or both.
/code-reviewContext preview
The summary Claude sees to decide when to auto-load this skill.
Review diffs and change sets for bugs, regressions, risks, and missing tests.
name: code-review
description: Review diffs and change sets for bugs, regressions, risks, and missing tests.
user-invocable: true
metadata:
hybridclaw:
category: development
short_description: "Diff and regression review."
tags:
- engineering
- review
- quality
- github
related_skills:
- github-pr-workflow
- feature-planningUse this skill to review local changes or an existing GitHub pull request for correctness, regressions, maintainability, security, and missing tests.
1. Establish scope with `git status --short`, `git diff --stat`, and `git diff --name-only`. 2. Read the changed files, not just the diff, before judging behavior. 3. Run targeted validation when it exists: tests, typecheck, lint, build, or repo-specific checks. 4. Present findings ordered by severity, then open questions, then a brief summary.
Use local git state when the change is in the current checkout:
git status --short git diff --stat git diff --name-only git diff --staged git diff <base-branch>...HEAD git log --oneline <base-branch>..HEAD
Replace `<base-branch>` with the repository's actual review base branch.
Use `gh` when the review target is an open pull request:
gh pr view 123 gh pr diff 123 gh pr checkout 123 gh pr view 123 --comments
After checking out the PR branch, review it the same way as a local branch and run the relevant repo checks before leaving comments.
Prioritize issues that change behavior or raise delivery risk:
Treat pure style comments as low priority unless the user explicitly asks for a style review.
Default to this structure:
1. Findings 2. Open questions or assumptions 3. Brief summary
For each finding:
If there are no findings, say that explicitly and mention any residual risk such as unrun integration tests or unverified deployment paths.
suggestions.
practical.
Use targeted searches when helpful:
git diff <base-branch>...HEAD | rg "console\\.log|TODO|FIXME|HACK|debugger" git diff <base-branch>...HEAD | rg "<<<<<<<|=======|>>>>>>>" git diff <base-branch>...HEAD | rg -i "password|secret|token|api[_-]?key|private[_-]?key"
Prefer repo-native test commands over broad guesses. Run the smallest check set that can confirm or disprove a suspected issue.
Enterprise-ready self-hosted AI assistant runtime with sandboxed execution, secure credentials, approvals, and memory
Install and use 1Password CLI (`op`) to sign in, inspect vault items, read secrets safely, and inject secrets into commands.
Coordinate cross-functional project delivery by tracking commitments, dependencies, decisions, and risks; maintain project…
Search Airtable bases and tables, read records and computed fields, and prepare guarded record CRUD requests with schema-based…
Expose HybridClaw as a custom Alexa skill and prepare guarded Alexa smart-home/device control payloads without exposing Amazon…
View Apple Calendar schedules, draft or import `.ics` files, and coordinate host-side calendar actions on macOS.
Control Apple Music playback, inspect now playing, start playlists, and automate the macOS Music app.