The Agent Skills Directory

[COMMAND_EXECUTION]: The skill constructs shell commands by interpolating user-supplied arguments such as version numbers, work item IDs, and dates into script execution strings (e.g., ./scripts/blueprint_release.sh <version> <work_item> <date>).- [REMOTE_CODE_EXECUTION]: The skill utilizes python3 -c to execute Python code that is dynamically generated. This includes inserting user-controlled strings into Python variable assignments and logic, which could be exploited for arbitrary code execution within the Python runtime environment.- [PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection due to the ingestion of untrusted user input that is then used directly in command templates.
Ingestion points: User-provided values for component names, version strings, and work item IDs found in SKILL.md instructions.
Boundary markers: None are defined to separate user data from the execution templates.
Capability inventory: Subprocess execution via the bash tool and file reading via the read tool.
Sanitization: There are no instructions or logic provided to validate or escape user inputs before they are interpolated into the shell and Python commands.
[EXTERNAL_DOWNLOADS]: The release process involves downloading build artifacts from the organization's Artifactory instance, which is a standard procedure for this type of release automation tool.

catalog-release