Feishu Multi-Agent Communication Tool

Problem: shared Feishu send targets across agents

When multiple agents need to send Feishu p2p or group messages under different identities, the pain is not only constructing the message payload. Sender settings can be scattered, targets can be selected incorrectly, and errors can be hard to trace. feishu-agent-send keeps sender identity, target agents, and p2p / group type in one local configuration so agents can share the same command conventions.

How it works: configure first, then output call instructions

It provides helper scripts: feishu_set_self.py for setting the agent's own chat_id, feishu_add.py for adding other agents, and feishu_who.py for listing available targets. For sending, prefer feishu_send.py --deliver, which outputs the full feishu_im_user_message call instruction and logging hint instead of requiring manual field assembly. If both private and group targets are configured, the tool prompts for switching, and --chat-type group reminds you to wait 6 seconds to reduce repeated sends. Error output includes error, hint, available_agents, and valid_options fields to make fixes easier.

Boundaries

It fits teams that already handle Feishu message calls and want consistent multi-agent sending behavior. It does not replace Feishu app permissions, callbacks, or card rendering. Missing sender info, unknown agents, or invalid chat_type values must be corrected using the provided hints.