The Agent Skills Directory

[PROMPT_INJECTION]: The skill uses natural instructional language to guide the AI's communication style (ASD-STE100 Simplified Technical English). There are no attempts to bypass safety filters or override system instructions.
[DATA_EXFILTRATION]: No evidence of credential exposure or data exfiltration. The skill references local documentation files for context but does not attempt to send data to external servers.
[COMMAND_EXECUTION]: No shell commands or script execution patterns were found. The skill explicitly disables model invocation, further reducing the attack surface.
[INDIRECT_PROMPT_INJECTION]:
Ingestion points: The skill reads ../../WORK-ROUTING.md and CONTEXT.md (found in SKILL.md).
Boundary markers: None present.
Capability inventory: No dangerous capabilities (file writes, network, or command execution) exist in the skill's code or instructions.
Sanitization: None present.

wait-what