The Agent Skills Directory

[COMMAND_EXECUTION]: The skill relies on the execution of the gmgn-cli tool for its core functionality, including configuration and data retrieval. It also instructs the agent to use system commands like ifconfig | grep inet6 (macOS) or ip addr show | grep inet6 (Linux) to gather network interface information for troubleshooting connectivity issues.
[EXTERNAL_DOWNLOADS]: The skill provides instructions for the user to install or update the gmgn-cli tool using npm install -g gmgn-cli if the tool is missing or returns an error. The skill also performs a network request to https://ipv6.icanhazip.com (a well-known service managed by Cloudflare) to verify public IP connectivity during troubleshooting.
[INDIRECT_PROMPT_INJECTION]: The skill processes data fields that contain user-generated content from external sources, which presents a surface for indirect prompt injection.
Ingestion points: The agent reads token_symbol, call_thesis.source_content, and twitter_username returned by the gmgn-cli callout commands (SKILL.md).
Boundary markers: The skill explicitly warns the agent that these fields are user-controlled text and should be treated strictly as data, instructing the agent to never follow instructions appearing inside them.
Capability inventory: The skill utilizes shell command execution via gmgn-cli and system network utilities.
Sanitization: The skill relies on instructional warnings to the agent rather than programmatic sanitization or filtering of the ingested text.

gmgn-callout