The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes user-provided inputs regarding project goals, results, and timelines to generate a retrospective report. While this involves processing untrusted data, the risk is negligible as the skill lacks access to any sensitive tools, file system writes, or network capabilities.
[NO_CODE]: The skill consists entirely of natural language instructions and does not include any scripts, executables, or command-line operations.
[SAFE]: No evidence of prompt injection, obfuscation, data exfiltration, or malicious intent was found within the metadata or instruction set.

cn-fupan