fresh-eyes-review
Skill1 starsChanged 8 months agoMITView on GitHub
Use when about to commit, create a PR, or declare work complete — after verification-before-completion confirms tests pass but before code ships. Catches security vulnerabilities, logic errors, and business rule bugs that slip through despite passing tests.
5 other projects use a version of this, 1 of them edited.
Adopted by
Most recently taken up by
- eai-org/agent-toolkittheir own version, first seen Oct 22, 2026
- soilSpoon/skillstheir own version, first seen Oct 20, 2026
- Dely0/fresh-eyes-reviewtheir own version, first seen Sep 16, 2026
- kcholakorn/fresh-eyes-reviewtheir own version, first seen May 29, 2026
- 2389-research/fresh-eyes-reviewedited, first seen Dec 5, 2025
Other versions5 projects use a version of this.
Word for word · 0
None found.
Edited · 1
- 2389-research/fresh-eyes-review★ 610% same
Wrote their own version · 4
- eai-org/agent-toolkit★ 85same name
- Dely0/fresh-eyes-review★ 2same name
- kcholakorn/fresh-eyes-review★ 1same name
- soilSpoon/skills★ 1same name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository5 other files it gives its agents.
- documentation-auditSkill
- frontend-designSkill
- landing-page-designSkill
- pythonSkill
- session-reflectionSkill
The file
Files
Fresh-Eyes Review
Overview
Deliberate re-reading of changed code with psychological distance. Catches what you assumed was correct.
Core principle: 100% test coverage can coexist with critical bugs.
Prerequisite: verification-before-completion must pass first.
Process
- Announce: "Starting fresh-eyes review of [N] files."
- Walk each changed file through the five checklists below.
- Fix immediately. Re-run tests after each fix.
- Declare: "Fresh-eyes complete. [N] issues found and fixed: [brief description of each]." Include this even for zero findings.
Checklists
Security
| Check | Look For |
| Injection | Unsanitized input in queries, commands, or templates |
| Path traversal | Unvalidated file paths, ../ sequences |
| Auth gaps | Unprotected endpoints, missing authorization checks |
| Secrets | Hardcoded credentials, tokens, or keys |
Logic
| Check | Look For |
| Boundaries | Off-by-one in indices, loops, pagination |
| Race conditions | Concurrent access to shared state |
| Null handling | Unguarded access chains that could throw |
| Error swallowing | Empty catch blocks, ignored rejections |
Business Rules
| Check | Look For |
| Calculations | Formulas matching requirements, correct rounding |
| Conditions | AND/OR logic correct, negations applied properly |
| Edge cases | Empty input, single item, zero, maximum values |
| Defaults | Sensible values when optional fields omitted |
Input Validation
| Check | Look For |
| Type checks | Expected types enforced at boundaries |
| Range checks | Numeric bounds, string lengths, array sizes |
| Format checks | Email, URL, date formats validated |
Performance
| Check | Look For |
| N+1 queries | Loops making individual database/API calls |
| Unbounded work | Missing limits on iterations, result sets, payloads |
| Resource leaks | Unclosed connections, streams, event listeners |
Resistance Patterns
| Rationalization | Reality |
| "Tests are comprehensive" | Tests validate design, not correctness |
| "I'm confident it's correct" | Confidence is inversely correlated with bugs |
| "It's just a small change" | Small changes cause large outages |
| "Partner is waiting" | 3 minutes now saves 3 hours debugging later |
| "Senior dev already approved" | They reviewed intent, not implementation details |
| "Production is blocked" | Rushing causes the outages being rushed to fix |
Red Flags — STOP
- "I already looked at this code while writing it"
- "The tests cover everything"
- "This is too trivial to review"
- "I just need to commit this quickly"
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.