The Agent Skills Directory

[SAFE]: The skill serves as a legitimate technical guide for the Latch bioinformatics platform. It contains no executable malicious code, obfuscation, or suspicious instructions.
[EXTERNAL_DOWNLOADS]: The documentation instructs users to install the official 'latch' Python package from standard registries and references documentation on the official 'latch.bio' domain. These are appropriate resources for the skill's purpose.
[CREDENTIALS_UNSAFE]: The skill demonstrates secure practices by using the get_secret function for handling sensitive API keys rather than hardcoding them or using unsafe storage methods.
[DATA_EXFILTRATION]: All data management operations described use the native 'latch:///' cloud storage protocol, which is restricted to the platform's workspace environment. No unauthorized data movement was detected.
[COMMAND_EXECUTION]: The shell commands provided (latch login, latch register, latch execute) are standard CLI operations for interacting with the Latch platform and are used appropriately within the tutorials.

latchbio-integration