Openclaw Skills for AgentMail (QQ) let you inspect schemas, send, read, list, and delete messages safely through the oo CLI.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install oo-agent-qq
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install oo-agent-qq using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is AgentMail (QQ) via Openclaw Skills?
Openclaw Skills for AgentMail (QQ) provide a secure, CLI-driven way to manage mailbox operations through an OOMOL-connected account. Instead of calling the API directly or handling raw credentials, the skill uses the agent_qq connector with the oo CLI, while OOMOL injects authentication server-side.
This skill is designed for dependable message automation with clear safety boundaries: read-only actions can run directly, while write and destructive actions require explicit user confirmation. Openclaw Skills also emphasize schema-first execution, so every payload is validated against the live connector contract before you send, update, or delete data.
AgentMail (QQ) via Openclaw Skills Use Cases
- Retrieve the live schema for an AgentMail (QQ) action before building a payload.
- List aliases and account limits for the current connected user.
- Search, enumerate, or inspect messages in an alias mailbox.
- Read a specific message body and attachment metadata.
- Send a message through AgentMail (QQ) with provider confirmation handled automatically when required.
- Move a message to trash with explicit approval for destructive operations.
- Recover from setup issues like missing CLI, expired credentials, or disconnected app connections.
- Use Openclaw Skills when you need safe, repeatable mailbox automation through the terminal.
How AgentMail (QQ) via Openclaw Skills Works
- Inspect the action contract with
oo connector schema "agent_qq" --action "<action_name>"to discover the authoritative input and output fields. - Build a JSON payload that matches the returned schema, or omit
--datato send an empty object when appropriate. - Execute the connector action with
oo connector run "agent_qq" --action "<action_name>" --data '<json>' --json. - Review the response payload, especially
meta.executionId, which identifies the execution for tracing and debugging. - For read-only actions, proceed directly since they are safe by default.
- For
[write]actions, confirm the exact payload and effect with the user before running. - For
[destructive]actions, obtain explicit approval for the target and outcome before executing. - If a command fails, run only the matching first-time setup step for the specific error condition, rather than repeating authentication or connection setup unnecessarily.
AgentMail (QQ) via Openclaw Skills Setup
- Install the oo CLI if it is not already available.
curl -fsSL https://cli.oomol.com/install.sh | bash
- Sign in to your OOMOL account if authentication is missing or invalid.
oo auth login
- Connect AgentMail (QQ) in the OOMOL console if the provider connection is missing, expired, or lacks required scopes.
https://console.oomol.com/app-connections?provider=agent_qq
- Refill billing only if you encounter a 402 or insufficient credit error.
https://console.oomol.com/billing/token-recharge
- Verify the connector schema before sending any payload to ensure your request matches the live contract.
oo connector schema "agent_qq" --action "list_messages"
- Run the desired action once the environment is ready.
oo connector run "agent_qq" --action "list_messages" --data '{"alias":"inbox"}' --json
AgentMail (QQ) via Openclaw Skills Data Schema & Taxonomy
Openclaw Skills for AgentMail (QQ) organize data around connector actions and their live schemas rather than a fixed local file format.
Core execution model
| Element | Description |
|---|---|
| Connector | agent_qq |
| CLI | oo |
| Payload source | Inline JSON string or @path/to/file.json |
| Default payload | {} if --data is omitted |
| Response shape | { "data": ..., "meta": { "executionId": "..." } } |
| Trace field | meta.executionId |
Available actions
| Action | Type | Purpose |
|---|---|---|
list_aliases |
read | Lists aliases and account limits |
list_messages |
read | Lists or searches messages in an alias mailbox |
read_message |
read | Returns message body and attachment metadata |
send_message |
write | Sends a message and may trigger provider confirmation |
delete_message |
destructive | Moves a message to trash and completes confirmation flow |
Metadata and safety taxonomy
- Untagged actions are read-only and can be executed directly.
[write]actions modify AgentMail (QQ) state and require user confirmation.[destructive]actions remove or overwrite data and require explicit approval.- Before building any payload, fetch the live schema with
oo connector schemaso your field names, types, and required properties stay aligned with the current provider contract.
AgentMail (QQ) via Openclaw Skills Advanced Features
- Schema-first execution with live contract inspection before every payload build.
- Automatic provider confirmation handling for supported send and delete workflows.
- Safety-aware action taxonomy that separates read, write, and destructive operations.
- No raw token handling because OOMOL injects credentials server-side.
- Supports both inline JSON payloads and file-based payloads via
@path/to/file.json. - Execution tracing through
meta.executionIdfor debugging and auditability. - Graceful first-time setup guidance for CLI, authentication, connectivity, and billing issues.
- Ideal for Openclaw Skills workflows that need repeatable, terminal-based AgentMail automation.
SKILL.md
Loading
Related Openclaw Skills
Daily Standup Notes
Openclaw Skills turns a messy brain dump into a polished daily standup update with Yesterday, Today, and Blockers sections.
LYGO Public Lattice Gate
LYGO Public Lattice Gate is a safe on-ramp skill for verifying public mirrors, measuring readiness, and generating dry-run presence proposals without live publishing.
My Notes Helper
My Notes Helper is an Openclaw Skills note-taking utility for storing quick references, daily snippets, and lightweight knowledge capture.
LYRA OpenClaw Hybrid
LYRA OpenClaw Hybrid is an operator-gated Openclaw Skills package that maps browser, Discord, memory, and launch limbs for secure hybrid automation.
Weekly Dev Tips
Weekly Dev Tips is a compact Openclaw Skills knowledge base that helps developers improve shell productivity, Git hygiene, and debugging judgment one actionable tip at a time.
find-ebooks / 搜好书
find-ebooks is an Openclaw Skills tool that searches ebooks across Anna's Archive and major Chinese book platforms, returning rich metadata and copyable links.
Featured*
