The Agent Skills Directory

[PROMPT_INJECTION]: The skill uses clear, directive language for task routing. No instructions were found that attempt to bypass safety filters, extract system prompts, or override core agent behavior.
[DATA_EXFILTRATION]: No sensitive file paths, hardcoded credentials, or network exfiltration patterns were detected. While the skill has access to network tools like WebFetch, their usage as described in the metadata is consistent with the primary analytics purpose.
[REMOTE_CODE_EXECUTION]: The skill does not perform any remote script downloads (e.g., curl-to-bash) or package installations at runtime.
[OBFUSCATION]: No obfuscated content, such as Base64-encoded strings, zero-width characters, or homoglyphs, was found in the skill definitions.
[PRIVILEGE_ESCALATION]: No attempts to escalate privileges, such as the use of sudo or modification of system configuration files, were identified.

lens