Analyze code diffs to derive change intent and assess review risk.
npx skills add https://github.com/markerikson/opencode-config-example --skill diffloupe
Please help me install this Agent Skill. Skill: diffloupe Source: https://github.com/markerikson/opencode-config-example/tree/main/config/skill/diffloupe Command: npx skills add https://github.com/markerikson/opencode-config-example --skill diffloupe
DiffLoupe helps you understand what a code diff changes, why it changed, and what risks it introduces so reviews don’t miss breaking, security, or intent mismatches.
Use diffloupe.ts analyze staged --json to review the intent, risks, and alignment of your currently staged changes in machine-readable form.
High-intent search queries and answers about installing and using this skill.
DiffLoupe analyzes code diffs to derive change intent and assess review risk across staged changes or PRs. It produces structured JSON output including intent, risk severity with evidence, and alignment matches or mismatches.
Semantic diff analysis for pull requests determines the likely purpose of code changes and evaluates hazards like bugs or security issues. It validates that actual modifications align with the stated intent from commit messages.
You validate stated commit intent by comparing provided text or commit messages against actual code modifications. DiffLoupe identifies intent alignment matches, mismatches, missing items, and unstated changes in structured JSON output.
Yes, you can review staged pre-commit changes by analyzing them for bugs, security issues, and breaking changes. DiffLoupe evaluates these hazards and provides risk severity with evidence tied to specific files.
DiffLoupe works with staged changes, pull requests, and commit ranges. It targets pre-commit review, PR understanding before merge, and validation that modifications align with stated intent across these input types.
The best way to assess review risk in unfamiliar code changes is deriving change intent and evaluating hazards. DiffLoupe produces structured output including risk severity with evidence, helping you understand breaking or security risks.
FabienDehopreGenerate Conventional Commits messages from git diffs and staged changes.
xpert-aiAssess dataset trustworthiness with severity ratings and root causes.
arcasilesgroupAutomate governed pull request workflows with pre-push gates and squash merge.
brenordvAudit software systems for security vulnerabilities and regulatory compliance.
