seo-ops
Warn
Audited by Gen Agent Trust Hub on Nov 17, 2026
Risk Level: MEDIUMCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
- [COMMAND_EXECUTION]: The
SKILL.mdpreamble contains shell commands (python3 telemetry/version_check.pyandpython3 telemetry/telemetry_init.py) that are executed automatically when the skill starts. The source code for these scripts is not provided within the skill package, which prevents verification of their behavior and intended purpose. - [INDIRECT_PROMPT_INJECTION]: The skill ingests data from several untrusted web sources and local files, creating a surface for potential injection attacks.
- Ingestion points:
trend_scout.pyfetches data from Google Trends RSS, Hacker News Firebase API, Reddit hot posts, and Brave Search results.content_attack_brief.pyreads local markdown and JSON files from a user-specified directory. - Boundary markers: Absent. The skill does not utilize protective delimiters or instructions to the agent to disregard potential commands embedded within the ingested data.
- Capability inventory: The skill possesses capabilities to write to the local filesystem (
./output), perform authenticated API requests to Google Search Console and Ahrefs, and open the system web browser (gsc_auth.py). - Sanitization: Absent. The ingested content is parsed and included in reports without filtering for executable instructions or prompt injection patterns.
- [DYNAMIC_EXECUTION]:
content_attack_brief.pyuses theimportliblibrary to dynamically load thegsc_client.pymodule at runtime rather than using standard static import statements. - [EXTERNAL_DOWNLOADS]: The skill performs legitimate network operations to fetch SEO data and trends from well-known services including Google Trends, Ahrefs, Reddit, Hacker News, and Brave Search.
Audit Metadata