The Agent Skills Directory

[NO_CODE] (SAFE): This skill is entirely instructional and does not include any scripts (.py, .js, .sh), binaries, or executables. All logic is handled by the AI agent via markdown instructions.
[PROMPT_INJECTION] (SAFE): No instructions were found that attempt to override system safety guidelines, extract system prompts, or bypass ethical constraints.
[DATA_EXFILTRATION] (SAFE): The skill does not perform any network operations (e.g., curl, fetch) or access sensitive local files (e.g., .env, .ssh).
[EXTERNAL_DOWNLOADS] (SAFE): No external scripts or packages are downloaded or executed.
[INDIRECT_PROMPT_INJECTION] (SAFE): While the skill processes untrusted user-provided copy, it lacks any dangerous capabilities (like file writing or network access) that could be exploited if malicious instructions were embedded in that copy.

copy-validator