The Agent Skills Directory

[PROMPT_INJECTION]: The skill identifies a potential attack surface for indirect prompt injection because it incorporates user-supplied threat objectives directly into logic-processing scripts.
Ingestion points: The DETECTION_OBJECTIVE variable in SKILL.md, which is populated via user input or environment variables.
Boundary markers: Absent; the user-provided string is processed without delimiters or instructions to ignore embedded commands.
Capability inventory: The agent can execute shell commands, perform file system operations in /tmp/, and install external packages.
Sanitization: No validation or escaping is performed on the objective string before it is passed to the Python analysis script via command-line arguments.
[COMMAND_EXECUTION]: The skill provides instructions for the execution of several CLI tools, including sigma-cli and python3, and suggests installing system dependencies using sudo apt install jq.
[EXTERNAL_DOWNLOADS]: The workflow involves downloading the sigma-cli package and SIEM-specific backend plugins from public registries and repositories.

siem-rule