The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user data, including company facts, metrics, and external news signals, to generate its output. While it lacks specific boundary markers (e.g., XML tags or delimiters) and explicit sanitization for this input, the risk is negligible because the skill does not request or utilize any dangerous capabilities such as file system writes, network access, or shell command execution. The instructions specifically emphasize grounding every claim in provided facts and refusing 'slop' or hallucinated content, which serves as a behavioral safety constraint.
[REMOTE_CODE_EXECUTION]: No remote code execution patterns, package installations, or dynamic code execution were detected. The skill relies entirely on natural language instructions for text generation and refers to other internal skills (e.g., journalist-fit-check) for workflow hand-offs rather than external scripts.

angle-generator