The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests data from external websites which could contain malicious instructions targeting the AI agent.\n
Ingestion points: The webfetch tool is used to pull content from /robots.txt, /sitemap.xml, and various faceted listing pages as described in the Data pulls section of SKILL.md.\n
Boundary markers: None. The instructions do not specify the use of delimiters or 'ignore' commands for the content fetched by webfetch.\n
Capability inventory: The skill executes Python scripts (scripts/dfs_client.py and scripts/report_publish.py) and generates reports based on the ingested data.\n
Sanitization: None. The process lacks validation or escaping steps for the external data before it is processed by the agent or the included scripts.\n- [COMMAND_EXECUTION]: The skill instructs the agent to execute specific local Python scripts to perform analysis and generate reports.\n
Evidence: python scripts/dfs_client.py onpage --url https://example.com --pretty\n
Evidence: python scripts/report_publish.py <report>.md

crawl-budget