The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes shell commands to execute playwright-cli for web interaction and runs a local Python script scripts/extract_page_signals.py to parse SEO and AEO signals. These actions are within the expected scope of the skill's functionality.
[PROMPT_INJECTION]: The skill presents an indirect prompt injection surface as it ingests untrusted data from third-party websites during analysis. This data is processed and presented to the agent for decision-making and report generation.
Ingestion points: SKILL.md (via content extraction from Playwright and the extract_page_signals.py script).
Boundary markers: No explicit boundary markers or 'ignore' instructions are provided to the agent when processing the extracted HTML or text content.
Capability inventory: The skill has access to sensitive tools (Google Search Console, Google Analytics) and can execute shell commands.
Sanitization: While the Python script extracts specific fields like headings and JSON-LD, it does not sanitize the text content (e.g., meta descriptions or paragraph text) for embedded instructions.
[EXTERNAL_DOWNLOADS]: The skill performs an initialization step using playwright-cli install --skills to download and configure browser binaries. This is a standard and safe operation for setting up the Playwright environment.

seo-analysis