Add fresh-eyes-review, review-squad, and simmer skills
@sophiealulachecks…add-2389-research-skills → master1 files · +3 −0updated 2mo agoloading diff…
@sophiealulachecks…add-2389-research-skills → master1 files · +3 −0updated 2mo ago@sophiealulachecks…add-2389-research-skills → master1 files · +3 −0updated 2mo agoThree Claude Code skills from 2389 Research to the Development & Code Tools section:
fresh-eyes-review — Pre-commit sanity check that catches security vulnerabilities, logic errors, and business rule bugs that pass tests. Runs through structured checklists (SQL injection, XSS, off-by-one, race conditions, etc.) before every commit.
review-squad — Dispatches parallel panels of specialized reviewer subagents. Includes default expert panels for web, API, mobile, and CLI projects (SEO, accessibility, security, performance, etc.) that run concurrently and consolidate into a severity-ranked report.
simmer — Iterative artifact refinement loop. Takes any artifact (code, docs, prompts, emails) and improves it over multiple rounds using investigation-first judges that actually read the code and propose evidence-based changes. Auto-selects single judge vs. judge board based on complexity.
We maintain ~25 skills at 2389 but picked these because they solve problems most Claude Code users hit:
All three are actively used in production across our team.
Added as external skill links in alphabetical order within Development & Code Tools, following the existing pattern for external skills (obra/superpowers, sanjay3290, NeoLabHQ, etc.).