Drop Soren into Claude and get a staff engineer who runs a real code review on whatever you paste in — bugs, security holes, performance traps, and maintainability debt — ranked by severity, with the fix shown as working code, not just described. Soren has reviewed thousands of pull requests at scale and has been paged at 3am for a bug that "looked fine in review" one too many times — he no longer gives anything a pass just because it compiles. He reads the whole submission before commenting on anything, hunts across four lenses in priority order, and never hands back vague feedback like "consider refactoring this" without saying exactly why and exactly what to type instead. What you get → One-line verdict up top — ship it or stop, in five seconds → Findings ranked Critical → High → Medium across correctness, security, performance, and maintainability → Every finding tied to a concrete failure scenario — the exact input, load, or sequence that triggers it → A corrected code snippet for every fix, not just a description of what to change → Security issues always called out explicitly as Critical — never buried as a style nit or inflated from a nitpick → Works across any common language or framework you paste in 📄 soren-code-auditor.skill Under 2 min install Works with Claude, ChatGPT & any AI chat How to install Download the .skill package → open Claude → paste SKILL.md into your Project Instructions or system prompt → paste in your code, the language/framework, and what it's supposed to do → Soren returns a severity-ranked report you could paste straight into a PR thread. Static read-through only — it doesn't execute or test the code, and isn't a substitute for automated SAST/DAST scanning.
