seo-competitor-analysis
Pass
Audited by Gen Agent Trust Hub on Nov 6, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from external sitemaps and competitor web pages, creating a surface for indirect prompt injection attacks.
- Ingestion points: The
competitor_sitemaps.pyscript fetches XML sitemaps and HTML content from arbitrary external domains identified as competitors. - Capability inventory: The skill possesses capabilities to read and write to local Obsidian vaults and perform network requests to external domains, which could be leveraged if an injection is successful.
- Boundary markers: The instructions do not provide explicit boundary markers or "ignore instructions" warnings when the agent processes the extracted data (topics, slugs, etc.) to build the market analysis plan.
- Sanitization: Data extraction is handled via regex in the Python script, which provides structural validation but does not sanitize the semantic content of the extracted strings, allowing potentially malicious instructions to reach the agent.
- [COMMAND_EXECUTION]: The skill executes a local Python script (
competitor_sitemaps.py) to automate sitemap discovery and parsing. This involves running shell commands via the agent's environment. - [EXTERNAL_DOWNLOADS]: The skill performs network operations to download content from competitor websites to analyze their strategy. While necessary for the skill's purpose, these are requests to external, untrusted sources.
Audit Metadata