The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes external data formats such as CSV, Excel, and JSON, which serves as an entry point for indirect prompt injection if the files contain hidden instructions.\n
Ingestion points: instructions.md specifies that the agent accepts project data and file paths; claw.json requests filesystem permissions.\n
Boundary markers: There are no explicit instructions or delimiters defined to prevent the agent from following instructions embedded within the data it processes.\n
Capability inventory: The agent has filesystem permissions and uses the pandas library, which can read and write data to the local environment.\n
Sanitization: The skill lacks mechanisms to sanitize or validate the content of the data files to ensure they do not contain malicious prompts.

df-merger