
from agents-to-im133
Manage and operate a Feishu/Lark bridge daemon for Claude Code/Codex: install, configure, start/stop, view logs, and run health checks.
This skill provides operational commands and documentation to manage a Feishu/Lark bridge daemon that connects Claude Code and Codex runtimes to IM groups. It includes setup guidance, CLI usage, runtime commands (onboard/start/stop/status/logs/doctor), and configuration examples.
Use this skill when deploying or operating an IM bridge for team collaboration: onboarding the bot, explaining required Feishu app configuration, starting/stopping the daemon, collecting logs, and running diagnostic checks. Do not attempt to start the daemon without a valid config.
{}
Agents-to-im is a Feishu/Lark bridge daemon for Claude Code and Codex, providing CLI commands for install, configure, start/stop, status, logs, and diagnostics. Scripts are well-written with proper shell safety (set -euo pipefail), platform-specific supervisors for macOS (launchd), Linux (setsid/nohup), and Windows (PowerShell/WinSW/NSSM). The dump-env.mjs script demonstrates thoughtful security design, preventing shell injection from config.env parsing. Doctor script ran and correctly reported missing config. Some scripts failed in the isolated test environment as expected — they require a real Feishu/Lark setup.
Well-structured ops skill for Feishu/Lark IM bridge. Security-conscious design: dump-env.mjs prevents shell injection from config parsing, logs redact sensitive values, no hardcoded secrets. External network calls to Feishu API are core functionality. eval usage in load_config_env is mitigated by the dump-env.mjs sanitization layer. Niche but real use case for Chinese market developers using Feishu.