The Agent Skills Directory

[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection (Category 8) as it ingests and processes untrusted data from user input or file content.
Ingestion points: User-provided text or file paths are captured in Step 1 and written to .galmuri/tmp/source-{slug}.txt.
Boundary markers: There are no instructions or delimiters specified to prevent the agent from following malicious commands embedded within the text being summarized.
Capability inventory: The skill utilizes the Bash tool for file management and the Skill tool to invoke an external summarization engine (galmuri:distill).
Sanitization: No sanitization, escaping, or validation of the input content is performed before processing.
[COMMAND_EXECUTION]: The skill employs the Bash tool to perform conditional logic and file operations.
Evidence: Step 1 uses shell commands to check for the existence of a warning file, touch new files, and pipe user input to temporary storage. While these specific operations are benign, they represent a surface for potential command injection if parameters like {slug} are not properly handled by the underlying platform.

explain