By 0101
Auto-generates review rules from instruction files and runs parallel code reviews
Investigates a single consolidated finding to determine whether it's real (Phase 3)
Consolidates and deduplicates findings from all Phase 1 discovery sources
Compiles the final review report and produces a user-facing summary
Reviews a diff chunk against a single review rule
Uses power tools
Uses Bash, Write, or Edit tools
Own this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimOwn this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimBased on adoption, maintenance, documentation, and repository signals. Not a security audit or endorsement.
A Copilot CLI / Claude Code plugin for thorough, consistent code reviews.
Focused Review breaks your review criteria into individual rules and concerns, then gives each one a dedicated context window. No competition for attention, no dropped instructions. The result is high recall and precision, even across dozens of review criteria.
A multi-phase pipeline then validates and refines findings:
The result: high-signal reviews with near-zero noise, enforcing your specific codebase standards across every PR.
CLAUDE.md, AGENTS.md, and other instruction files and extracts review criteria as individual rulesreview/ as Markdown, reviewed in PRs like codeSome review criteria need broad context and benefit from multiple perspectives — things like architectural coherence, error handling patterns, or security posture. These get expressed as concerns: higher-level review prompts that are run by multiple LLM models in parallel. Where rules are precise and narrow, concerns cast a wider net across your codebase.
This separation also lets you scale cost and quality independently. Simple style rules run on fast, cheap models with small focused contexts. Deep concerns like security get multiple frontier models that spend time understanding the codebase for a proper in-depth review.
You choose which rules and concerns to enable, and at what scale — from a lightweight pass with a few built-in defaults to a comprehensive review with dozens of project-specific rules and multi-model concerns.
copilot plugin marketplace add 0101/focused-review
copilot plugin install focused-review@focused-review
/plugin marketplace add 0101/focused-review
/plugin install focused-review@focused-review
The skill is conversational — you can ask it to help you set things up:
/focused-review:review Help me set this up!
Or use the built-in commands to configure step by step:
/focused-review:review configure # set rules directory, discover instruction files
/focused-review:review refresh # generate rules from your instruction files
refresh scans your instruction files (CLAUDE.md, AGENTS.md, .github/copilot-instructions.md, etc.), extracts review criteria, and writes them as individual rule files to review/. If no instruction files are found, it copies the built-in defaults — general review, code duplication, plus security, bugs, and architecture concerns.
Review the generated rules in review/, enable or disable what you need, and commit them. From here on, rules are version-controlled like any other code.
Run a focused review against the current branch diff (default):
/focused-review:review
Specify a diff scope:
/focused-review:review branch # diff against origin/main (default)
/focused-review:review commit # diff of last commit
/focused-review:review staged # staged changes only
/focused-review:review unstaged # unstaged changes only
/focused-review:review full # scan entire codebase (no diff)
Rules directory is controlled by a focused-review.json config file:
{ "rules_dir": "custom-rules/" }
The config file is discovered from these locations (first match wins):
.claude/focused-review.json -- project shared (version-controlled)focused-review.json -- repo root.github/focused-review.json -- GitHub convention~/.claude/focused-review.json -- user-wide (Claude Code)~/.copilot/focused-review.json -- user-wide (Copilot CLI)If no config file is found, defaults to review/.
Resolution priority: explicit --rules-dir CLI flag > focused-review.json config file > review/ default.
Run /focused-review:review configure to create or update the config file interactively.
npx claudepluginhub 0101/focused-review --plugin focused-reviewMulti-model consensus engine integrating OpenAI Codex CLI, Gemini CLI, and Claude CLI for collaborative code review and problem-solving.
Ultra-compressed communication mode. Cuts ~75% of tokens while keeping full technical accuracy by speaking like a caveman.
Comprehensive UI/UX design plugin for mobile (iOS, Android, React Native) and web applications with design systems, accessibility, and modern patterns
Curate auto-memory, promote learnings to CLAUDE.md and rules, extract proven patterns into reusable skills.
Harness-native ECC operator layer - 67 agents, 271 skills, 92 legacy command shims, reusable hooks, rules, selective install profiles, and production-ready workflows for Claude Code, Codex, OpenCode, Cursor, and related agent harnesses