dojo.md
University for AI agents. 92 courses, 4400+ scenarios, any model via OpenRouter. Auto-training loops generate per-model SKILL.md documents. Works with Claude Code, OpenClaw, Cursor, Windsurf. No fine-tuning req...
Based on FeiShu(飞书) / Lark's OpenAPI MCP server, manage user information, chats, emails, cloud documents, multidimensional tables, tasks, calendars, etc.
Suivez les instructions du dépôt pour votre agent. Consultez les fichiers et dépendances avant l’installation.
SkillHub CLI
skillhub install al-one/mcp-larkExaminez la source et la compatibilité, puis suivez les étapes d’installation.
Donnez le lien à votre agent et demandez-lui d’examiner instructions et dépendances avant d’expliquer les étapes.
Need to login to the Lark MCP Configuration Platform to add MCP services, obtain the MCP URL, and configure it into environment variables.
Prioritize fetching from .env under the workspace, then use system environment variables. If not configured, ask the user for input and update it to .env.
# Configure multiple MCP service URLs and usage scenarios in environment variables
LARK_MCP_SERVERS='
open.larksuite.com/mcp/stream/xxx:Chats and Mails;
open.larksuite.com/mcp/stream/yyy:Cloud documents;
'
npx -y mcporter list 'open.larksuite.com/mcp/stream/<token>' --all-parameters
# Get the schema of the specified tool
npx -y mcporter list 'open.larksuite.com/mcp/stream/<token>' --json | jq '.tools[] | select(.name == "<tool_name>")'
npx -y mcporter call 'open.larksuite.com/mcp/stream/<token>.<tool_name>' param1:"value1" foo:"bar"
mcporterTo improve compatibility, use npx -y mcporter instead of mcporter when executing commands.
Ces informations proviennent de l’analyse indexée et peuvent être anciennes. Vérifiez les fichiers actuels.
# MCP Lark / FeiShu Need to login to the Lark MCP Configuration Platform to add MCP services, obtain the MCP URL, and configure it into environment variables. - Lark MCP docs: https://open.larksuite.com/document/mcp_open_tools/call-feishu-mcp-server-in-remote-mode - 飞书 MCP 文档: https://open.feishu.cn/document/mcp_open_tools/end-user-call-remote-mcp-server ## Environment variables Prioritize fetching from `.env` under the workspace, then use system environment variables. If not configured, ask the user for input and update it to `.env`. ```shell # Configure multiple MCP service URLs and usage scenarios in environment variables LARK_MCP_SERVERS=' open.larksuite.com/mcp/stream/xxx:Chats and Mails; open.larksuite.com/mcp/stream/yyy:Cloud documents; ' ``` ## List of available tools ```shell npx -y mcporter list 'open.larksuite.com/mcp/stream/<token>' --all-parameters # Get the schema of the specified tool npx -y mcporter list 'open.larksuite.com/mcp/stream/<token>' --json | jq '.tools[] | select(.name == "<tool_name>")' ``` ## Call specified tool ```shell npx -y mcporter call 'open.larksuite.com/mcp/stream/<token>.<tool_name>' param1:"value1" foo:"bar" ``` ## References - Sent message content: references/message_create.md ## About `mcporter` To improve compatibility, use `npx -y mcporter` instead of `mcporter` when executing commands.
University for AI agents. 92 courses, 4400+ scenarios, any model via OpenRouter. Auto-training loops generate per-model SKILL.md documents. Works with Claude Code, OpenClaw, Cursor, Windsurf. No fine-tuning req...
macOS WPS Office workflow helper skill for safer document preparation, conversion, export, and compatibility guidance
A股分析技能
A2A4B2B Skill for OpenClaw