The Agent Skills Directory

[SAFE] (SAFE): The skill primarily consists of documentation and conversion strategies. All referenced actions (Bash, Read, Write) are consistent with the intended purpose of auditing local skill files and generating conversion outputs.
[EXTERNAL_DOWNLOADS] (SAFE): No remote code downloads or external network operations were identified. The conversion logic for 'GPT Actions' is described conceptually and does not include active network triggers.
[COMMAND_EXECUTION] (SAFE): While Bash is listed in allowed-tools, it is intended for local file management during the conversion audit. No dangerous or obfuscated command-line patterns were found.
[PROMPT_INJECTION] (SAFE): No attempts to override system instructions or bypass safety filters were detected. The instructions are focused on the task of formatting and condensing information for platform compatibility.
[INDIRECT_PROMPT_INJECTION] (LOW): The skill processes other SKILL.md files as input.
Ingestion points: Reads SKILL.md and scripts from user-specified directories.
Boundary markers: absent (uses standard read tools).
Capability inventory: Bash, Read, Write tools available for local file operations.
Sanitization: absent.
Reasoning: While the skill reads external content, it does so to perform a conversion task. The risk of unintended instruction execution from the source file is low and consistent with the tool's primary purpose.

openai-gpt-converter