The Agent Skills Directory

[PROMPT_INJECTION]: The skill defines a protocol for ingesting content from external images and PDFs via the look-at utility, creating an indirect prompt injection surface where the agent processes and acts upon untrusted data. (1) Ingestion points: External media files processed by look_at.py as described in SKILL.md. (2) Boundary markers: The skill does not define specific delimiters or instructions to ignore embedded content within the extracted media data. (3) Capability inventory: The agent has access to file system operations, tool execution via Bash, and task delegation via the Task tool. (4) Sanitization: No evidence of sanitization or filtering is provided for the output of the media extraction tool.
[COMMAND_EXECUTION]: The skill utilizes uv run python3 to execute local scripts like look_at.py and skill-metadata.py within the plugin root for operational tasks.
[EXTERNAL_DOWNLOADS]: References the obra/superpowers repository on GitHub as a source for development and behavioral patterns.

using-skills