The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted content from various manufacturer and trade websites via search and fetch tools to generate recommendations and update the local product library. This creates a vulnerability surface where malicious instructions embedded in external web data could attempt to influence the agent's logic.
Ingestion points: Technical specifications, product descriptions, and search result snippets retrieved from the web in SKILL.md (Step 2).
Boundary markers: The instructions do not specify the use of delimiters or specific instructions to ignore embedded commands within the fetched data.
Capability inventory: The skill uses Bash, Write, and Edit to update the product-library.csv file within the project directory.
Sanitization: There are no explicit instructions for sanitizing or escaping content retrieved from manufacturer websites before it is serialized and written to the project library.
[EXTERNAL_DOWNLOADS]: The skill uses WebFetch to retrieve data from external URLs to gather purchasing facts and technical specifications during the research process.
[COMMAND_EXECUTION]: The skill utilizes the Bash tool for project navigation and managing file system operations related to appending research data to local CSV files.

product-research