The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill utilizes npx ruv-swarm throughout its documentation and workflows. This command downloads and executes code from the npm registry at runtime. The source organization (ruvnet) is not a recognized trusted vendor, which introduces supply chain risks.
[COMMAND_EXECUTION]: The provided webhook-handler.js example demonstrates a critical security flaw by using execSync with unsanitized string interpolation from event.comment.body. An attacker could exploit this by posting a pull request comment designed to execute arbitrary shell commands on the server hosting the webhook.
[DYNAMIC_EXECUTION]: The skill allows for the registration of custom review agents from local files via npx ruv-swarm github register-agent --file "./custom-review-agent.js". This capability allows for the dynamic loading and execution of arbitrary JavaScript code.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process untrusted data from GitHub Pull Requests, including code diffs, descriptions, and comments, which could contain malicious instructions meant to manipulate the review agents or the automated workflow.
Ingestion points: GitHub PR data retrieved via gh pr view and webhook payloads.
Boundary markers: None identified; data is passed directly as arguments to tools.
Capability inventory: Execution of shell commands via gh and npx, modifying repository state, and posting comments.
Sanitization: The skill documentation lacks sanitization or validation logic for external inputs, particularly in the webhook integration examples.

qe-github-code-review