mirror of
https://github.com/pocketpaw/pocketpaw.git
synced 2026-05-19 00:17:08 +00:00
* feat(cli): add 10 new subcommands for local agent management New CLI commands: - doctor: full diagnostics with connectivity checks - health: quick startup-only health check (no network) - channels: list channel status, start/stop adapters via REST - skills: list and search available skills - sessions: list, delete, and search chat sessions - memory: show stats and search long-term memories - config: show (masked), set, validate, path - errors: show recent errors with --limit and --search - logs: show/tail audit log with --follow All commands support --json for scripting. Refactored __main__.py to subcommand-based parser while keeping full backwards compat with existing flags (--doctor, --telegram, --discord, etc). * docs: add new CLI subcommands to CLAUDE.md * docs: add CLI reference page with all subcommands New docs page at getting-started/cli covering all 15+ CLI commands with usage examples, flags, and descriptions. Added to sidebar nav after Quick Start.
245 lines
5.1 KiB
Plaintext
245 lines
5.1 KiB
Plaintext
---
|
|
title: "CLI Reference: All PocketPaw Commands"
|
|
description: "Complete reference for the PocketPaw CLI. Manage your local AI agent from the terminal: health checks, channels, skills, sessions, memory, config, errors, logs, and more."
|
|
section: Getting Started
|
|
ogType: article
|
|
keywords: ["cli", "commands", "terminal", "pocketpaw cli", "command line", "subcommands"]
|
|
tags: ["cli", "reference", "getting-started"]
|
|
---
|
|
|
|
# CLI Reference
|
|
|
|
PocketPaw ships with a full set of CLI commands for managing your local AI agent without the web dashboard. Every command supports `--json` for scripting and automation.
|
|
|
|
## Starting PocketPaw
|
|
|
|
```bash
|
|
# Web dashboard (default, opens browser)
|
|
pocketpaw
|
|
|
|
# API-only server (no dashboard UI)
|
|
pocketpaw serve
|
|
|
|
# Development mode (auto-reload on file changes)
|
|
pocketpaw --dev
|
|
|
|
# Custom host and port
|
|
pocketpaw --host 0.0.0.0 --port 9000
|
|
```
|
|
|
|
## Headless Channel Modes
|
|
|
|
Run PocketPaw as a headless bot on one or more channels:
|
|
|
|
```bash
|
|
pocketpaw --telegram # Telegram-only mode
|
|
pocketpaw --discord # Discord bot
|
|
pocketpaw --slack # Slack bot (Socket Mode)
|
|
pocketpaw --whatsapp # WhatsApp webhook server
|
|
pocketpaw --discord --slack # Multiple channels simultaneously
|
|
```
|
|
|
|
## Agent Status
|
|
|
|
```bash
|
|
# Show current agent status
|
|
pocketpaw status
|
|
|
|
# JSON output for scripting
|
|
pocketpaw status --json
|
|
|
|
# Live monitoring (refresh every 2 seconds)
|
|
pocketpaw status --watch
|
|
|
|
# Custom refresh interval
|
|
pocketpaw status --watch 5
|
|
```
|
|
|
|
## Health and Diagnostics
|
|
|
|
```bash
|
|
# Quick health check (startup checks only, no network calls)
|
|
pocketpaw health
|
|
|
|
# Full diagnostics (config + connectivity + version check)
|
|
pocketpaw doctor
|
|
|
|
# JSON output
|
|
pocketpaw health --json
|
|
pocketpaw doctor --json
|
|
```
|
|
|
|
`health` runs fast, offline checks (config exists, API key format, disk space, etc.). `doctor` additionally tests network connectivity to your configured LLM provider.
|
|
|
|
## Channels
|
|
|
|
```bash
|
|
# List all channels with configured/autostart status
|
|
pocketpaw channels
|
|
|
|
# Start a channel adapter (requires a running PocketPaw instance)
|
|
pocketpaw channels start discord
|
|
|
|
# Stop a channel adapter
|
|
pocketpaw channels stop slack
|
|
|
|
# JSON output
|
|
pocketpaw channels --json
|
|
```
|
|
|
|
<Callout type="info">
|
|
`channels start` and `channels stop` communicate with a running PocketPaw instance via its REST API. The dashboard or `pocketpaw serve` must be running first.
|
|
</Callout>
|
|
|
|
## Skills
|
|
|
|
```bash
|
|
# List all available skills
|
|
pocketpaw skills
|
|
|
|
# JSON output
|
|
pocketpaw skills --json
|
|
```
|
|
|
|
Skills are loaded from three directories (in priority order):
|
|
- `~/.agents/skills/` (AgentSkills ecosystem)
|
|
- `~/.claude/skills/` (Claude Code / SDK standard)
|
|
- `~/.pocketpaw/skills/` (PocketPaw-specific)
|
|
|
|
## Sessions
|
|
|
|
```bash
|
|
# List recent chat sessions
|
|
pocketpaw sessions
|
|
|
|
# Search session content
|
|
pocketpaw sessions search "deployment script"
|
|
|
|
# Delete a session
|
|
pocketpaw sessions delete <session-key>
|
|
|
|
# Limit results
|
|
pocketpaw sessions --limit 50
|
|
|
|
# JSON output
|
|
pocketpaw sessions --json
|
|
```
|
|
|
|
## Memory
|
|
|
|
```bash
|
|
# Show memory stats (long-term entries, daily notes, session count)
|
|
pocketpaw memory
|
|
|
|
# Search long-term memories
|
|
pocketpaw memory search "API integration"
|
|
|
|
# Limit search results
|
|
pocketpaw memory search "project" --limit 20
|
|
|
|
# JSON output
|
|
pocketpaw memory --json
|
|
```
|
|
|
|
## Configuration
|
|
|
|
```bash
|
|
# Show current config (sensitive values masked)
|
|
pocketpaw config
|
|
|
|
# Set a config value
|
|
pocketpaw config set agent_backend openai_agents
|
|
pocketpaw config set web_port 9000
|
|
|
|
# Validate API keys and settings
|
|
pocketpaw config validate
|
|
|
|
# Print config file path
|
|
pocketpaw config path
|
|
|
|
# JSON output
|
|
pocketpaw config --json
|
|
pocketpaw config validate --json
|
|
```
|
|
|
|
Config values are automatically coerced to the correct type (boolean, integer, list, etc.).
|
|
|
|
## Errors
|
|
|
|
```bash
|
|
# Show last 20 errors
|
|
pocketpaw errors
|
|
|
|
# Show more errors
|
|
pocketpaw errors --limit 50
|
|
|
|
# Filter errors
|
|
pocketpaw errors --search "timeout"
|
|
|
|
# JSON output
|
|
pocketpaw errors --json
|
|
```
|
|
|
|
Errors are sourced from the health engine's persistent error store.
|
|
|
|
## Audit Logs
|
|
|
|
```bash
|
|
# Show last 50 audit log entries
|
|
pocketpaw logs
|
|
|
|
# Tail the audit log in real time
|
|
pocketpaw logs --follow
|
|
|
|
# Show fewer entries
|
|
pocketpaw logs --limit 10
|
|
|
|
# JSON output
|
|
pocketpaw logs --json
|
|
```
|
|
|
|
The audit log is stored at `~/.pocketpaw/audit.jsonl`.
|
|
|
|
## Provider Diagnostics
|
|
|
|
```bash
|
|
# Test Ollama connectivity, model availability, and tool calling
|
|
pocketpaw --check-ollama
|
|
|
|
# Test OpenAI-compatible endpoint
|
|
pocketpaw --check-openai-compatible
|
|
```
|
|
|
|
## Security
|
|
|
|
```bash
|
|
# Run security audit
|
|
pocketpaw --security-audit
|
|
|
|
# Auto-fix issues found by audit
|
|
pocketpaw --security-audit --fix
|
|
|
|
# Scan memory files for PII
|
|
pocketpaw --pii-scan
|
|
```
|
|
|
|
## Update
|
|
|
|
```bash
|
|
# Check for updates and install via uv
|
|
pocketpaw update
|
|
```
|
|
|
|
## Global Flags
|
|
|
|
| Flag | Description |
|
|
|---|---|
|
|
| `--json` | Output as JSON (works with most subcommands) |
|
|
| `--limit N` | Limit number of results |
|
|
| `--follow` | Tail mode (for `logs`) |
|
|
| `--watch [N]` | Live refresh every N seconds (for `status`, default: 2) |
|
|
| `--host HOST` | Bind web server to specific host |
|
|
| `--port PORT` | Web server port (default: 8888) |
|
|
| `--dev` | Development mode with auto-reload |
|
|
| `--version` | Show version |
|