The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses a local Python script scripts/check_plain_language.py to validate that non-technical specifications do not contain banned technical terminology. The script is benign and uses standard libraries for regex-based text analysis.
[INDIRECT_PROMPT_INJECTION]: The skill processes user-provided drafts, attachments, and external documentation as part of its core workflow. To mitigate risks, it includes an explicit 'Step 0' for intent recognition and guardrails to identify and ignore attempts to reveal system prompts or bypass rules. It also enforces the use of code blocks for intermediate data to maintain clear boundaries between processing steps.

spec-organizer