The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes a bash command engine orchestrate next --scope $ARGUMENTS using the {{INVOKE}} directive. This invokes a local binary or script named engine with arguments provided at runtime.
[INDIRECT_PROMPT_INJECTION]: The skill serves as a surface for injection by interpolating untrusted input into a sensitive execution context.
Ingestion points: User-supplied input via the $ARGUMENTS variable in SKILL.md.
Boundary markers: None. The skill does not use delimiters or provide instructions to the agent to ignore potentially malicious embedded content.
Capability inventory: Execution of shell commands via bash as defined in the skill's logic.
Sanitization: None. The instructions explicitly state to "pass every argument through verbatim," which bypasses standard shell escaping and allows for command injection if metacharacters like ;, &, or | are included in the input.

aidlc-scope