multi-agent-feishu 多飞书账号接入不同agent
ReviewAudited by ClawScan on May 1, 2026.
Overview
This looks like a straightforward setup guide for connecting multiple Feishu bot accounts to separate OpenClaw agents, but users should handle the bot secrets, message routing, and persistent agent memory carefully.
Before installing or following this skill, confirm that you control the Feishu apps being added, protect the OpenClaw configuration and backups because they contain app secrets, verify each agent-to-bot binding, and remember that each agent may retain its own conversation history and state.
Findings (4)
Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.
Running the setup can change which agent receives and replies to messages for a Feishu bot account.
This command creates a new OpenClaw agent, assigns a workspace, and binds it to a Feishu account. It is user-directed and aligned with the purpose, but it changes gateway behavior and should be reviewed before running.
openclaw agents add <agent_id> --workspace ~/.openclaw/workspace<N> --bind feishu:<account_id> --non-interactive
Run the commands manually, verify each agent ID, workspace, and Feishu account ID, and inspect the resulting bindings before using the bots in real chats.
Anyone who can read the configuration file may be able to use or abuse the configured Feishu bot credentials.
The skill requires Feishu app credentials to be stored in the OpenClaw configuration. This is expected for a Feishu bot integration, but app secrets are sensitive account credentials.
"<account_id>": { "appId": "cli_xxx", "appSecret": "xxx" }Use only Feishu apps you control, grant the minimum needed permissions, protect ~/.openclaw/openclaw.json and backups with appropriate file permissions, and rotate secrets if they are exposed.
Messages, memories, and skill settings may remain available to the same agent later, which can be helpful but may also retain sensitive or stale information.
The architecture explicitly creates persistent per-agent memory, session history, and skill configuration. This supports the stated isolation goal, but it means chat context and agent state may persist across tasks.
每个 Agent 有独立的：... **记忆** (`memory/`) - 独立的会话历史 ... **技能** (`skills/`) - 独立的技能配置
Keep workspaces separated, periodically review or clear agent memory when needed, and avoid placing highly sensitive conversations in bots unless retention is acceptable.
A mistaken binding could send a Feishu conversation to the wrong agent or cause replies from the wrong bot identity.
The documented data flow routes Feishu chat messages through the OpenClaw Gateway to selected agents and sends replies back through the configured bot account. This is expected for the integration, but users should confirm routing boundaries.
用户在飞书@机器人发送消息 ... 飞书将消息推送到 Gateway ... Gateway 根据 `bindings` 匹配到对应的 Agent ... 通过对应机器人的飞书账号发送回复
Check `openclaw agents list --bindings`, keep bot membership limited to intended chats, and test each account-agent binding before production use.