The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from external API responses, creating a surface for Indirect Prompt Injection.
Ingestion points: Data enters the context through response.json() and response.text as described in Step 2 and Step 3 of SKILL.md.
Boundary markers: The instructions do not specify the use of delimiters (like XML tags or distinct markers) to isolate API response content from the agent's internal reasoning or subsequent instructions.
Capability inventory: The skill utilizes curl for shell-based requests and the Python requests library for network operations, providing a functional path for automated actions based on input.
Sanitization: While the guidelines recommend masking sensitive values in final reports, there is no explicit instruction to sanitize or escape data before interpolating it into the agent's prompt context.
[COMMAND_EXECUTION]: The skill instructions provide templates for executing curl commands in a shell environment and Python scripts using the requests library. These are used to interact with user-defined URLs, which is the primary intended function of the skill.
[EXTERNAL_DOWNLOADS]: The skill performs network operations to fetch data from external URLs provided by the user or discovered during API testing. This is standard behavior for an API testing tool.

api-tester