SIDDARTHA577/envguard-node / .agents/skills/envguard-code-review
envguard-code-review
UnverifiedConducts rigorous, evidence-based code reviews evaluating correctness, edge cases, performance bottlenecks, architectural alignment, test adequacy, security best practices, and breaking changes.
envguard-code-review Skill
Guides engineers and AI agents through conducting rigorous, evidence-based code reviews of pull requests, commits, and feature branches. Replaces generic feedback with structured findings across logic correctness, error handling, security boundaries, performance efficiency, and test adequacy.
When to Trigger
- Reviewing pull requests, git diffs, or newly implemented files before merging to protected branches.
- Inspecting code changes for logical bugs, unhandled promise rejections, race conditions, or off-by-one errors.
- Evaluating database queries, loops, and async patterns for N+1 performance bottlenecks or memory leaks.
- Auditing code modifications for security regressions (SQL injection, XSS, prototype pollution, secret logging).
- Assessing automated test adequacy and determining whether boundary conditions and edge cases are covered.
- Detecting breaking API changes, contract divergence, or backward compatibility risks in shared modules.
When NOT to Trigger
- Drafting initial system architecture requirements from scratch (use
envguard-system-design). - Running automated stylistic formatting or lint fixing without qualitative review (use standard
eslint --fixorprettier). - Auditing environment configuration files and deployment secrets exclusively (use
envguard-revieworenvguard-audit). - Conducting automated accessibility audits on UI components (use
envguard-accessibility).
Supported Projects & Environments
- Languages: TypeScript, JavaScript (Node.js, Browser, ESM, CommonJS), HTML, CSS, SQL.
- Source Control: Git, GitHub PRs, GitLab MRs, local staged/unstaged changes.
- Frameworks & Libraries: Express, React, Vite, Next.js, Fastify, Vitest, Jest.
- Development Stacks: Full-stack Node.js/TypeScript applications, CLI tools, libraries.
Required Tools & Permissions
- Tools: File reader, command runner, diff inspector (
read_file,run_command). - Permissions: Project workspace read access to review source files, diffs, test logs, and configuration. Read-only by default; write permissions only if requested to generate test patches.
Expected Inputs
- Git diff, pull request summary, or list of modified source files.
- Problem statement or feature requirements describing the intended behavior.
- Existing test suite results and linting reports.
Step-by-Step Workflow
Step 1: Context & Change Scope Analysis
- Inspect the diff or commit range using
git diffor file inspection. - Determine the scope: Is this a bug fix, feature addition, refactoring, dependency upgrade, or performance patch?
- Check whether the changes adhere to project conventions, architectural guidelines, and scope boundaries. Flag unrelated edits or scope creep.
Step 2: Correctness & Logic Verification
- Trace data flow from inputs to outputs:
- Are null, undefined, and empty array/object inputs handled safely?
- Are asynchronous operations properly awaited (
Promise.allvs sequentialawaitin loops)? - Are there race conditions in shared mutable state or module-level variables?
- Are array indices, loop termination conditions, and boundary values correct (e.g.
<vs<=)?
- Verify that return types and error statuses match API contracts.
Step 3: Error & Exception Handling Audit
- Verify that every async function has proper try/catch or error propagation.
- Ensure promises do not fail silently or lead to unhandled rejections.
- Verify that error handlers log actionable diagnostics without leaking internal stack traces, system paths, or database credentials to end users.
- Check that resources (file descriptors, database transactions, network streams) are cleaned up in
finallyblocks.
Step 4: Security & Sensitive Data Boundary Check
- Check for common vulnerabilities:
- Injection: Ensure SQL queries use parameterized placeholders (
$1,?), never string interpolation. - Prototype Pollution: Verify object mergers and deep clones guard against
__proto__andconstructor. - ReDoS: Check that regular expressions do not contain nested quantifiers evaluated on untrusted input.
- Secret Leakage: Ensure
console.log()or telemetry does not output passwords, authorization headers, or environment secrets. - Authorization: Ensure all protected routes verify user identity and tenancy permissions.
- Injection: Ensure SQL queries use parameterized placeholders (
Step 5: Performance & Resource Efficiency Review
- Detect algorithmic inefficiencies:
- Avoid N+1 database queries in loops; verify batching or eager loading with
IN (...)or JOINs. - Inspect module-level caches: Do they have size limits or TTLs, or will they cause memory leaks over time?
- Verify streaming for large file uploads/downloads rather than buffering entire payloads into memory.
- Check for unnecessary re-renders in React components (missing dependencies in
useEffect/useMemo).
- Avoid N+1 database queries in loops; verify batching or eager loading with
Step 6: Test Adequacy Assessment
- Inspect test files associated with the change:
- Do tests exercise the happy path?
- Do tests exercise edge cases: invalid inputs, empty sets, network failures, unauthorized access?
- Are mocks realistic, or do they bypass the logic being tested?
- Do tests assert specific results rather than simply verifying that code executed without throwing?
Step 7: Formulate Structured Review Report
Organize findings using the standard EnvGuard Review Taxonomy:
- Blocker / Critical: Serious logic bug, data corruption risk, severe security vulnerability, or build failure. Must be resolved before merge.
- Major / High: Unhandled edge case, memory leak, missing transaction rollback, or breaking API contract. Strongly recommended before merge.
- Minor / Medium: Sub-optimal performance pattern, code duplication, or incomplete test coverage.
- Nit / Suggestion: Code readability improvement, docstring clarification, or minor naming refinement. Non-blocking.
Validation Steps
- Execute
npm run lintandnpm run typecheckto confirm static analysis passes cleanly. - Execute
npm testto confirm all existing and new unit/integration tests pass. - Verify that every reported finding includes a concrete file path, line reference, problem explanation, and suggested fix.
Expected Outputs
- Structured Markdown Code Review Report containing:
- Summary of changes.
- Blocker & Major findings with code snippets and remediation steps.
- Minor improvements and suggestions.
- Test adequacy evaluation and test coverage recommendations.
- Merge recommendation (
Approve,Request Changes, orComment).
Failure Handling
- If the change cannot be understood without additional business context, clearly state the missing assumptions and ask focused clarifying questions rather than guessing.
- If automated tests fail in CI, prioritize reporting the root cause of the test failure before reviewing stylistic aspects.
Safety & Privacy Boundaries
- Never paste proprietary business logic or real production customer data into unvetted external services.
- Never approve code changes that disable authentication, bypass EnvGuard schema validation, or turn off security linters.
Example User Prompts
- "Perform a rigorous code review of pull request #12 with envguard-code-review"
- "Review our new transaction processing service for race conditions, error handling, and memory leaks"
- "Audit these recent TypeScript changes for potential breaking API changes and test coverage gaps"
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| rea-analysis-optimizationmorluto/reaOptimize REA providers and shared analysis workflows for agent context, latency, CPU, and memory while preserving correctness and Evidence. Use for analysis bottlenecks and cache or query design; CI/test-suite optimization belongs to rea-ci-tests-audit. | Stars: 464k | Updated | Claude CodeCodex | Verified | Install |
| rea-ci-tests-auditmorluto/reaAudit REA CI and test suites for correctness, fixture drift, false passes, and measured performance; use for suite or workflow investigations and efficiency changes, rather than ordinary feature testing or a single test failure. | Stars: 464k | Updated | Claude CodeCodex | Verified | Install |
| rea-tool-designmorluto/reaAudit, investigate, design, or change REA investigation tools, CLI/MCP contracts, provider capabilities, and Evidence semantics. | Stars: 464k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-auditDietrichGebert/ponytailQuality audit of the whole repo: bugs, security, real load, missing tests, speed, and what to delete. Most important first. | Stars: 192k | Updated | Claude CodeCodex | Verified | Install |