The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes shell commands (curl, ffuf) and shell script loops to perform network enumeration and vulnerability testing against user-defined targets as part of its penetration testing functions.
[REMOTE_CODE_EXECUTION]: The skill provides static Python script templates that involve network requests and file system operations. These scripts are intended for execution by the agent to automate data extraction from targets during authorized testing.
[PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface by ingesting and processing data from external web targets.
Ingestion points: Data is retrieved from target application endpoints (e.g., in SKILL.md, the Step 6 Python script fetches data from https://TARGET/api/users/{}/profile).
Boundary markers: The skill does not employ boundary markers or instructions to ignore embedded commands when the agent processes response data from the target.
Capability inventory: The skill possesses network access capabilities (via curl, ffuf, and Python's requests library) and file writing capabilities (via Python's json.dump to idor-dump.json).
Sanitization: There is no evidence of sanitization, validation, or filtering of the external data before it is written to the local filesystem.

idor