The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructions direct the agent to execute shell commands (curlCommand) provided as output from the testany_list_credential_safes and testany_list_credential_keys tools. Executing dynamically generated strings as shell commands is a high-risk pattern.
[EXTERNAL_DOWNLOADS]: The agent is instructed to use curl to interact with dynamic network endpoints (TSSM gateway) to retrieve metadata for credentials based on tool output.
[REMOTE_CODE_EXECUTION]: The skill manages the upload and execution of script packages (ZIP files) via a 'Dry Run' mechanism. This involves downloading or processing code in various languages (Python, Java, etc.) and executing it within a runtime environment.
[PROMPT_INJECTION]: The skill possesses an attack surface for indirect prompt injection via the processing of untrusted external data.
Ingestion points: Ingests 'platform case packages' and 'test scenarios' from upstream sources or users (SKILL.md).
Boundary markers: Absent; the instructions do not specify the use of delimiters or warnings to ignore instructions embedded within the ingested script packages or scenario descriptions.
Capability inventory: The skill has access to network operations (curl) and significant platform-side CRUD capabilities.
Sanitization: No validation or sanitization logic is mentioned for the external data processed by the agent.

testany-case