astro-seo
Warn
Audited by Gen Agent Trust Hub on Oct 8, 2026
Risk Level: MEDIUMEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONCREDENTIALS_UNSAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill documentation encourages the AI agent to suggest the installation of companion skills from an external GitHub repository (coreyhaines31/marketingskills) that is not part of the trusted vendor list.
- [REMOTE_CODE_EXECUTION]: Instructions guide the agent to prompt users to execute
npx skills addusing URLs from a third-party GitHub account, facilitating the download and execution of external code. - [COMMAND_EXECUTION]: The skill provides scripts for full and incremental IndexNow submission intended for use in CI/CD pipelines. These scripts perform recursive directory scanning and outbound network requests.
- [CREDENTIALS_UNSAFE]: The IndexNow submission process transmits a sensitive API key to an external service. While standard secret management is recommended, the submission endpoint is configurable via environment variables, which could allow the key to be sent to an arbitrary host if the environment is compromised.
- [INDIRECT_PROMPT_INJECTION]: The skill presents an attack surface for indirect prompt injection by processing external data and user-provided strings without robust isolation.
- Ingestion points: Component props in
JsonLd.astroand input received via stdin or CLI arguments in the IndexNow scripts. - Boundary markers: The skill lacks explicit delimiters or instructions to ignore embedded commands within processed data.
- Capability inventory: The included Node.js scripts can read local files and perform network operations.
- Sanitization: Sanitization is minimal; for example,
JsonLd.astrouses standard JSON stringification which may not prevent script injection if input strings contain closing script tags.
Audit Metadata