The Agent Skills Directory

[DATA_EXFILTRATION]: The skill functionality rns_status accesses the ~/.reticulum directory. This path is highly sensitive as it is the default storage location for Reticulum network identity keys and private configuration files.
[INDIRECT_PROMPT_INJECTION]: The skill includes a fetch_page tool and references external 'nomadnet_context' prompts, which allows for the ingestion of untrusted remote content into the agent context.
Ingestion points: External text retrieved via fetch_page as described in SKILL.md.
Boundary markers: No delimiters or instructions are present to ensure the agent ignores malicious instructions embedded in the fetched content.
Capability inventory: The skill has the ability to read local sensitive filesystem state and perform documentation searches.
Sanitization: There is no evidence of validation, escaping, or sanitization for the remote content before it is processed.

rns