The Agent Skills Directory

[PROMPT_INJECTION]: The skill processes user-supplied input through the $ARGUMENTS variable to perform question mirroring. This represents a surface for indirect prompt injection where malicious instructions in the input could influence the model.
Ingestion points: SKILL.md (Step 1)
Boundary markers: Absent. The skill does not use delimiters to wrap the injected arguments.
Capability inventory: No dangerous capabilities (file writes, network requests, or subprocess execution) are defined within the skill files.
Sanitization: Absent. The input is mirrored directly without validation.
[EXTERNAL_DOWNLOADS]: The documentation provides links to the author's repository for installation and packaged versions of the skill.
Evidence: References to github.com/synapseradio/ai-skills in README.md for npx installation and .skill file downloads.

ask-respond