Perform a deep evidence-backed PsychSift repository audit and produce severity-ranked, reproducible findings without mutating code unless fixes are requested. Use for repo-wide audits, risk assessments, or exhaustive review requests.
84%
Does it follow best practices?
Run evals on this skill
Adds up to 20 points to the overall score
Low
Low-risk findings worth noting
The skill exposes the agent to untrusted, user-generated content from public third-party sources, creating a risk of indirect prompt injection. This includes browsing arbitrary URLs, reading social media posts or forum comments, and analyzing content from unknown websites.
The required workflow in step 1 reads PR comments for a `Review record` on any pull request being audited, and GitHub PR comments can be posted by outsiders, so the agent ingests outsider-authored free text without first selecting a specific item.
BigSimmo/PsychSift
0d5a97a
If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.