The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill instructs the user or agent to clone a repository from an external source (https://github.com/23blocks-OS/ai-maestro-plugins.git) to obtain the required messaging scripts.
[COMMAND_EXECUTION]: The installation process involves executing a shell script (install-messaging.sh -y) downloaded from the external repository, which allows for the execution of arbitrary remote code on the host system.
[INDIRECT_PROMPT_INJECTION]: The core functionality of the skill involves reading and processing messages from other agents via amp-inbox and amp-read, which could contain malicious instructions designed to influence the receiving agent's behavior.
Ingestion points: Message bodies and subjects retrieved from the Agent Messaging Protocol (AMP) identity inbox.
Boundary markers: The instructions do not specify any delimiters or safety markers to isolate incoming message content from the agent's system instructions.
Capability inventory: The skill allows the agent to send messages (amp-send), reply to messages (amp-reply), delete records (amp-delete), and manage identities (amp-init).
Sanitization: There is no mention of sanitization, filtering, or validation performed on the contents of the messages before they are presented to the agent.

agent-messaging