The Agent Skills Directory

[SAFE]: The skill provides structured instructions for a multi-phase writing process, including data retrieval, parallel writing using subagents, and final assembly.
[COMMAND_EXECUTION]: The skill uses restricted Bash commands (mkdir, mktemp, date) for standard filesystem and environment management tasks.
[DATA_EXFILTRATION]: All data operations are confined to the local project directory, reading from outline.md and research/sources.json and writing to article.md and run-config.json.
[PROMPT_INJECTION]: The procedural instructions are focused on formatting and citation rules without any attempt to bypass safety filters or override agent constraints.

storm-write