The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses unzip and ls commands to manage the diagnostic dump files. This is a standard operation for the skill's purpose but involves shell interaction with external files.
[DATA_EXFILTRATION]: While the skill accesses and reports on sensitive configuration data like agentId and platform URLs (anypoint.mulesoft.com), these actions are limited to the intended diagnostic report and the resources belong to the legitimate vendor domain.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it processes untrusted content from the ZIP file and JSON resources without explicit boundary markers or sanitization logic.
Ingestion points: dump.zip, api-instances.json, policy-bindings.json, services.json, configurations.json and user-provided conf.d/ directories.
Boundary markers: No specific delimiters or instructions to ignore embedded prompts are provided in the analysis steps.
Capability inventory: The agent has the ability to execute shell commands (unzip, ls) and read arbitrary files from the extracted directory.
Sanitization: No data validation or escaping is performed on the content of the JSON fields before they are processed by the agent.

omni-gateway-dump