joeholdcroft/code-review-feedback
Systematically handle GitHub PR review feedback: fetch comments, plan responses, make code changes, and reply to reviewers with explicit approval at each stage.
72
90%
Does it follow best practices?
Run evals on this skill
Adds up to 20 points to the overall score
View guide
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 user-generated GitHub content (see Phase 2: the GraphQL query that retrieves reviewThreads and reviews and the `gh pr view --json comments` call), and the workflow requires parsing those untrusted review/comment bodies to decide actions (Phase 3–8) such as code changes, replies, and resolving threads, so third-party content can directly influence agent behavior.
Where we found it
github.com (implicit, via `gh api graphql` and `gh pr view` CLI calls)
domain · 5 sites
The plugin fetches PR metadata, review threads, review bodies, and PR-level comments from GitHub's API via `gh pr view` (line 25) and multiple `gh api graphql` calls (lines 37, 84, 90, 227), then parses and acts on that untrusted third-party content; the domain "github.com" is the implicit target of all these calls, not stated literally in the source.
SKILL.md
25
gh pr view --json number,url,title,headRefName
SKILL.md
37
gh api graphql
SKILL.md
84
gh api graphql
SKILL.md
90
gh api graphql
SKILL.md
227
gh api graphql
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 uses the gh CLI to fetch PR review content at runtime from GitHub (e.g., https://api.github.com/graphql and PR URLs like https://github.com/OWNER/REPO/pull/NUMBER), and those fetched review bodies/comments are parsed and injected into the agent’s processing to generate plans and replies, so external content directly controls the agent's instructions and is a required dependency.
Where we found it
gh
dependency · 11 sites
`gh` is explicitly declared as a prerequisite the user must install and authenticate independently (line 13: "The `gh` CLI must be authenticated"); the plugin invokes it as a local system tool, not as a remotely fetched or installed dependency, so it does not meet the W012 definition.
- Audited
- Security analysis