seo-weekly-report
Pass
Audited by Gen Agent Trust Hub on Oct 24, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data, such as search queries, page titles, and referring URLs, from Google Search Console, Google Analytics 4, and Bing APIs to generate a natural language Markdown report.
- Ingestion points: Data enters the agent's context through API clients in
fetch-gsc.mjs,fetch-ga4.mjs, andfetch-bing.mjs. - Boundary markers: The report generation instructions in
references/portable/report-template.mddo not utilize explicit delimiters or isolation techniques to separate raw API data from the agent's instructions. - Capability inventory: The skill facilitates file system writes (
reports/*.md) and broad network access to search engine APIs. - Sanitization: While the scripts perform basic text normalization for brand-matching logic, there is no evidence of sanitization designed to prevent the agent from interpreting the extracted data as instructions.
- [COMMAND_EXECUTION]: The orchestration script
scripts/fetch-all.mjsutilizesspawnSyncto manage the execution of internal sub-scripts. - The commands are strictly limited to the local directory using
__dirnameandprocess.execPath, preventing the execution of arbitrary external commands. - [EXTERNAL_DOWNLOADS]: The skill relies on standard and well-known dependencies from the npm registry, including
googleapis,@google-analytics/data, anddotenv. - These are official libraries from trusted technology organizations and are used for their intended analytical purposes.
Audit Metadata