Install
$ agentstack add skill-sylphai-inc-atskills-code-review ✓ scanned · ✓ verified, works with Claude Code, Cursor, and more.
Security review
✓ PassedNo issues found. Passed automated security review. · v0.1.0 How review works →
- ✓ Prompt-injection patterns
- ✓ Secret / credential exfiltration
- ✓ Dangerous shell & filesystem operations
- ✓ Untrusted network calls
- ✓ Known-malicious package signatures
What it can access
- ✓ Network access No
- ✓ Filesystem access No
- ✓ Shell / process execution No
- ✓ Environment & secrets No
- ✓ Dynamic code execution No
From automated source analysis of v0.1.0. “Used” means the capability is present in the source — more access means more to trust, not that it’s unsafe.
Verified badge
Passed review? Show it. Paste this badge into your README, it links to the public security report.
Reliability & compatibility
Declared compatibility
Compatibility is declared by the source manifest. End-to-end runtime verification is coming, see below.
We're building live execution health for every listing: tool-call success rate, median latency, uptime, and last-checked timestamps, measured, not self-reported. It isn't live yet, so we don't show numbers we can't stand behind.
How agent discovery & health will work →About
Code Review Workflow
Review the given diff, pull request, or set of changed files as a careful senior engineer would — looking for correctness, safety, and maintainability issues, not just style nits.
Steps
- Understand intent first. Read the PR description / commit messages / linked issue before reading the diff line-by-line. Know what problem the change is supposed to solve before judging whether it solves it.
- Trace the change end-to-end. For each modified function/module, identify every caller and every downstream consumer of its output. A change that looks correct in isolation can still break a caller that assumed the old behavior.
- Check correctness before style. In priority order:
- Logic errors — off-by-one, incorrect conditionals, wrong operator, inverted boolean logic.
- Error handling — are error/exception paths handled, or silently swallowed? Are edge cases (empty input, null/None, zero, max values) covered?
- Concurrency/state — race conditions, shared mutable state, missing locks/transactions where needed.
- Security — unsanitized input reaching a query/shell/template, secrets in code, missing auth checks on new endpoints.
- Resource handling — unclosed files/connections, unbounded loops/memory growth, missing timeouts.
- Check test coverage. Does the diff include tests for the new/changed behavior? Do the tests actually exercise the changed logic, or do they just assert trivial properties (tautological tests)? Are edge cases from step 3 covered?
- Check for duplication and reuse. Does this diff reimplement something that already exists elsewhere in the codebase (a date formatter, a retry helper, an API client)? Prefer reusing/extending existing utilities over introducing parallel implementations.
- Check the blast radius. Is the change scoped to what the task requires, or does it touch unrelated files/formatting/comments? Flag unrelated changes explicitly — they hide the real diff and increase review risk.
- Only then, style and readability. Naming, comments, consistency with surrounding code conventions. Note these, but don't let them dominate the review — they're the last priority, not the first.
- Write the review as specific, actionable comments. For each issue found:
- Point to the exact file/line.
- State what's wrong (not just "this looks off") and why it matters (what could break, and under what condition).
- Suggest a concrete fix or ask a clarifying question if intent is genuinely ambiguous.
- Distinguish blocking issues ("must fix before merge") from non-blocking suggestions ("consider," "nit:").
Output Format
Summarize as:
- TL;DR — does this change look safe to merge, with or without required fixes?
- Blocking issues — must be resolved before merge, with file/line references.
- Non-blocking suggestions — nice-to-haves, style, minor improvements.
- Questions — anything genuinely ambiguous that needs the author's input before you can finish the review.
Source & license
This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.
- Author: SylphAI-Inc
- Source: SylphAI-Inc/atskills
- License: MIT
- Homepage: https://atskills.one
Install and usage instructions live in the source repository linked above.
Reviews
No reviews yet, be the first.
Write a review
Versions
- v0.1.0 Imported from the upstream source.