The Agent Skills Directory

[SAFE]: The skill defines a configuration for a multi-agent messaging system using local CLI tools. No malicious patterns, obfuscation, or unauthorized exfiltration attempts were detected.
[COMMAND_EXECUTION]: The skill utilizes the ccgram command-line interface to manage the agent swarm. This includes ccgram msg spawn for creating new agent instances with specific prompts, which is a standard feature of the described multi-agent orchestration workflow.
[PROMPT_INJECTION]: The skill processes messages from other agents in the swarm, creating a surface for potential indirect prompt injection. However, the skill provides clear safety guidelines, instructing the agent to summarize incoming messages for the user and request explicit confirmation before executing tasks requested by peers.

ccgram-messaging