sahildmk/pr-comment-resolver
Review PR comments, address code issues in source files (not generated files), regenerate derived artifacts, run lint/format, commit, push, and reply to the comment thread confirming resolution.
92
90%
Does it follow best practices?
Impact
97%
1.21xAverage score across 5 eval scenarios
Low
Low-risk findings worth noting
Low
Low-risk findings.
2 low severity findings. Worth noting, but not necessarily harmful.
W011: Third-party content exposure detected (indirect prompt injection risk).
What this means
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.
Why it was flagged
The skill explicitly fetches and reads user-generated PR content via the "Identify the PR and fetch PR feedback" workflow (e.g., `gh api repos/{owner}/{repo}/pulls/{pr_number}/comments`, issues comments, and reviews) and uses those third-party review bodies to decide edits, commits, and replies, so untrusted reviewer text could inject instructions that affect the agent's actions.
Where we found it
github.com
domain · 6 sites
The plugin fetches PR review comments, issue comments, and review summaries from GitHub's API; these contain untrusted third-party text (reviewer/bot feedback) that the agent reads and acts upon to decide code edits, commits, and replies.
W012: Unverifiable external dependency detected (runtime URL that controls agent).
What this means
The skill fetches instructions or code from an external URL at runtime, and the fetched content directly controls the agent’s prompts or executes code. This dynamic dependency allows the external source to modify the agent’s behavior without any changes to the skill itself.
Why it was flagged
The skill issues runtime GitHub API calls (e.g., "gh api repos/{owner}/{repo}/pulls/{pr_number}/comments") to fetch reviewer comments that are ingested as instructions to the agent and thus remote content can directly control agent behavior.
Where we found it
gh
command · 12 sites
The plugin instructs the agent to execute the `gh` CLI at runtime to make GitHub API calls that fetch remote content (PR comments, reviews) which is then ingested as instructions for the agent.
- Audited
- Security analysis