The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data which introduces a surface for indirect prompt injection.
Ingestion points: Step 1 in SKILL.md instructs the agent to read ticket.md, research documents, and any files mentioned by the user or within those artifacts.
Boundary markers: The instructions do not mandate the use of delimiters or 'ignore' warnings for the external content being processed.
Capability inventory: The skill possesses the ability to list directories (ls -La), query git metadata (git rev-parse HEAD), and write markdown files to the .agent-work/tasks/ directory.
Sanitization: No specific sanitization or filtering of the ingested external content is described.
[COMMAND_EXECUTION]: The skill utilizes standard command-line tools for directory enumeration and metadata collection.
Evidence: SKILL.md uses ls -La to verify task directories, and references/design_discussion_template.md specifies the use of git rev-parse HEAD to populate document frontmatter. These operations are benign and restricted to project-level observation.

create-design-discussion