The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructions provide shell commands that directly interpolate user-supplied data into command-line arguments.
Evidence: In SKILL.md, the Data pulls and Output sections specify commands such as python scripts/dfs_client.py ideas --keyword "<pillar-topic>" and python scripts/report_publish.py <report>.md.
Risk: User-controlled strings for topics or filenames could contain shell metacharacters (e.g., ;, &&, or |), which could lead to unauthorized command execution if not properly escaped by the platform.
[INDIRECT_PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection because it processes untrusted data from users and external tools without sufficient safeguards.
Ingestion points: Untrusted data enters the agent context through the pillar topic, candidate spoke topics, and the results returned by scripts/dfs_client.py.
Boundary markers: There are no boundary markers or instructions to the agent to treat external content as data rather than instructions.
Capability inventory: The skill utilizes shell execution for Python scripts and file system writes for markdown reports.
Sanitization: No sanitization, validation, or escaping logic is described for the data before it is interpolated into shell commands or content plans.

supporting-content-planner