Part of Run AI-Assisted Development Workflows End to End
Engineering Blueprint
Audit Code Quality and Find Bugs Before Landing
Specialist-driven pre-landing review that catches what CI doesn't (architecture, security, edge cases).
What problem does this solve?
Reviews your diff for structural issues that tests miss: SQL safety, race conditions, LLM output trust boundaries, enum completeness, and more. /review runs a critical-pass checklist against your branch, dispatches specialist subagents (testing, maintainability, security, performance, data migration, API contract, design), and aggregates findings. It pairs with /codex for a second opinion from OpenAI's model. Because reviews happen pre-push, you fix issues while they're still fresh.
How does it work?
- When your branch is ready, run
/review(or gstack invokes it during/ship). 2. It runs the critical-pass checklist, dispatches 6-8 specialist subagents in parallel (testing, security, performance, maintainability, etc.). 3. Each specialist audits the diff and flags findings by confidence level. 4. Findings are merged, deduplicated, and presented for auto-fix (LOW and INFORMATIONAL) or user approval (CRITICAL and HIGH). 5. Auto-fixed items are committed. 6. User approves or skips remaining findings. Output: clean diffs with known issues traced.
Included Skills
No skills in this group
What's required to run this?
Reviews read the diff against the base branch (git diff origin/<base>...HEAD). Specialist subagents run via the Agent tool (parallel dispatch). Greptile integration requires a GitHub PR to exist. Results log to ~/.gstack/projects/{slug}/$BRANCH-reviews.jsonl for cross-session tracking. Plan completion audit cross-references plan items against the diff.
Key Features
/review
Pre-landing review with critical-pass checklist (SQL safety, race conditions, LLM boundaries) plus 6-8 specialist subagents that run in parallel.
/codex
Independent code review from OpenAI Codex CLI. Three modes: pass/fail gate, adversarial challenge, open consultation. Compares with `/review` for cross-model synthesis.
/design-review
Designer's-eye audit of UI/UX in production. Rates design on six dimensions, auto-fixes slop (typos, spacing), screenshots before/after.
/devex-review
Live developer experience audit for APIs, CLIs, SDKs. Navigates docs, times TTHW, tests getting-started flows, compares against `/plan-devex-review` plan.
Greptile Integration
Resolves Greptile review comments (when using GitHub integration) with reply templates and evidence-backed reasoning.
About This Blueprint
- Industry
- Technology
More Blueprints to explore
Business Automation & Productivity Workflow
Reduce hours lost to repetitive admin without adding a complicated stack. Prioritized, tested workflows automate routine research, client communication, project tracking, and website tasks while keeping people in control of important decisions.
Business Technology & Digital Solutions Workflow
Avoid costly, fragmented digital projects with a plan that connects business goals to practical tools, implementation steps, and measurable results. It also flags performance, security, accessibility, and SEO checks before launch.
ICP-Filtered Outreach Lists from G2 Buyer Intent
Stop wasting rep time on weak intent signals and mismatched accounts. The workflow ranks ICP-fit buyers by readiness, finds the right contact, and supplies tailored outreach angles for the strongest opportunities.
ClaudeG2 Buyer Intent (G2 MCP Server)ZoomInfo (MCP)+3 moreIshani D.
Account Manager at G2