You can run it by following steps:
- Prepare the configuration file
config.yaml. An example configuration file is provided atconfig.yaml.example:
# Bot configuration
app_id: <app_id> # or set it from cli options `--app-id`
app_secret: <app_secret> # or set it from cli options `--app-secret`
# Cherry pick configuration
cherry_pick_invite:
audit_webhook: "https://open.feishu.cn/open-apis/bot/v2/hook/xxx"
github_token: "ghp_xxx"
# Ask command configuration
ask:
llm:
azure_config:
api_key: "your-api-key"
base_url: "https://your-deployment.openai.azure.com"
api_version: "2023-05-15"
model: "gpt-4"
system_prompt: "You are a helpful assistant for PingCAP employees."
mcp_servers:
tidb_server:
base_url: "https://mcp-server.example.com"
# DevBuild configuration
devbuild:
api_url: "https://tibuild.pingcap.net/api/devbuilds"
# RepoAdmins command configuration
repo_admins:
audit_webhook: "https://open.feishu.cn/open-apis/bot/v2/hook/xxx"
github_token: "ghp_xxx"- Run the lark bot app:
go run ./cmd/server [-app-id=<your_app_id>] [-app-secret=<your_app_secret>]We are deploying the bot with GitOps using FluxCD, the manifest is in the chatops-lark directory in ee-ops.
So if you want to bump the image or update the configuration, you contribute a PR to the repository.