CLI Reference

The Actionbook CLI (actionbook) is the primary interface for interacting with Actionbook. Every browser command is stateless — pass --session and --tab explicitly.

`actionbook search`

Search for available action manuals. Usage:

actionbook search [query]

Arguments:

query: The search keyword (e.g., “youtube”, “youtube upload”).

Example:

actionbook search "youtube"

`actionbook get`

Retrieve a specific action manual by its ID. Usage:

actionbook get [actionId]

Arguments:

actionId: The unique identifier of the action.

Example:

actionbook get "site/airbnb.com/page/home/element/search-button"

`actionbook manual`

Get detailed manual information for a site, group, or action. Alias: man. Usage:

actionbook manual [site] [group] [action]

Arguments:

site: Site name (e.g., “youtube”, “airbnb”). If omitted, lists available sites.
group: Group name within the site (e.g., “videos”).
action: Action name within the group (e.g., “search”).

Examples:

actionbook manual youtube                         # Overview of a site
actionbook manual youtube videos                  # Actions in a group
actionbook manual youtube videos search           # Detailed action docs
actionbook man youtube --json                     # Alias with JSON output

`actionbook setup`

Initial setup wizard for API key, browser preferences, health checks, and optional skills install. Usage:

actionbook setup
actionbook setup --target claude          # Quick mode: install skills for an agent
actionbook setup --non-interactive --api-key <KEY>
actionbook setup --reset

Targets: claude, codex, cursor, windsurf, antigravity, opencode, hermes, standalone, all

`actionbook browser`

Browser automation commands.

After actionbook setup, browser command syntax is identical across all modes (local, cloud, extension).

Global Flags

These flags apply to all actionbook browser subcommands:

Flag	Description
`--session <id>`	Target a specific session (defaults to `default`)
`--tab <id>`	Target a specific tab
`--json`	Output results as JSON envelope
`--timeout <ms>`	Set command timeout in milliseconds

Session Lifecycle

`actionbook browser start`

Start or attach a browser session. Entry point for all modes including cloud providers. Options:

Flag	Description
`--mode <mode>`	Browser mode: `local`, `extension`, or `cloud`
`-p, --provider <name>`	Cloud browser provider: `driver`, `hyperbrowser`, `browseruse`. Implies `--mode cloud`
`--session <id>`	Session ID (get-or-create: reuse if exists, create if not)
`--set-session-id <id>`	Session ID (get-or-create: same as `--session`)
`--open-url <url>`	Open this URL on start
`--cdp-endpoint <url>`	Connect to an existing CDP endpoint
`--header <KEY:VALUE>`	Headers for CDP endpoint (may be repeated)
`--headless`	Run in headless mode
`--profile <name>`	Profile name
`--stealth` / `--no-stealth`	Anti-detection mode (default: enabled)
`--max-tracked-requests <N>`	Network request buffer size per tab (default: 500, range: 1–100000)

Examples:

actionbook browser start --set-session-id s1
actionbook browser start --session s1 --open-url https://google.com
actionbook browser start -p hyperbrowser --session s1
actionbook browser start --mode cloud --cdp-endpoint wss://browser.example.com/ws

-p is mutually exclusive with --cdp-endpoint and --mode local/extension. See Cloud Providers for provider-specific env vars.

Both --session and --set-session-id are get-or-create: they reuse a Running session with the given ID, or create one if not found. When reusing, if --profile is passed and does not match the session’s bound profile, the command fails with SESSION_PROFILE_MISMATCH:

error.code: "SESSION_PROFILE_MISMATCH" (retryable: false)
error.hint: "omit --profile or pass --profile <bound> to reuse"
error.details: { session_id, bound_profile, requested_profile }

Omitting --profile or passing a matching value allows reuse.

Other Session Commands

actionbook browser list-sessions                      # List all active sessions
actionbook browser status --session s1                # Show session status
actionbook browser close --session s1                 # Close a session (alias: stop)
actionbook browser restart --session s1               # Restart, preserving session_id

browser close is idempotent: closing an unknown or already-closed session returns ok: true with meta.warnings instead of a fatal error. This prevents false alarms in orchestrators that call close unconditionally during cleanup.

Unknown/already-closed session: ok: true, data: { status: "closed", closed_tabs: 0 }, meta.warnings: ["session not found in daemon — already closed or daemon restarted"]
Another close already in flight: SESSION_CLOSING fatal (unchanged)

Tab Management

actionbook browser list-tabs --session s1                         # List tabs
actionbook browser new-tab "https://example.com" --session s1     # Open new tab (alias: open)
actionbook browser close-tab --session s1 --tab t1                # Close a tab

actionbook browser goto <url> --session s1 --tab t1              # Navigate to URL
actionbook browser back --session s1 --tab t1                    # Go back
actionbook browser forward --session s1 --tab t1                 # Go forward
actionbook browser reload --session s1 --tab t1                  # Reload page

goto supports --wait-until to control when the command returns: domcontentloaded (default), load, or none.

Observation

# Snapshot
actionbook browser snapshot --session s1 --tab t1                # Accessibility tree with refs
actionbook browser snapshot -i --session s1 --tab t1             # Interactive elements only
actionbook browser snapshot -i -c --session s1 --tab t1          # Interactive + compact

# Page info
actionbook browser title --session s1 --tab t1                   # Page title
actionbook browser url --session s1 --tab t1                     # Current URL
actionbook browser viewport --session s1 --tab t1                # Viewport dimensions

# Content
actionbook browser text --session s1 --tab t1                    # Full page text
actionbook browser text "<selector>" --session s1 --tab t1       # Element text
actionbook browser html --session s1 --tab t1                    # Full page HTML
actionbook browser html "<selector>" --session s1 --tab t1       # Element HTML
actionbook browser value "<selector>" --session s1 --tab t1      # Input value

# Element inspection
actionbook browser attr "<selector>" href --session s1 --tab t1        # Single attribute
actionbook browser attrs "<selector>" --session s1 --tab t1            # All attributes
actionbook browser box "<selector>" --session s1 --tab t1              # Bounding rect
actionbook browser styles "<selector>" color fontSize --session s1 --tab t1  # Computed styles
actionbook browser describe "<selector>" --session s1 --tab t1         # Full element description
actionbook browser state "<selector>" --session s1 --tab t1            # State flags
actionbook browser inspect-point 420,310 --session s1 --tab t1         # Inspect at coordinates

# Query
actionbook browser query one "<selector>" --session s1 --tab t1       # Exactly one match
actionbook browser query all "<selector>" --session s1 --tab t1       # All matches
actionbook browser query count "<selector>" --session s1 --tab t1     # Match count
actionbook browser query nth 2 "<selector>" --session s1 --tab t1     # Nth match (1-based)

# Screenshot & PDF
actionbook browser screenshot output.png --session s1 --tab t1        # Screenshot
actionbook browser screenshot output.png --full --session s1 --tab t1  # Full page
actionbook browser screenshot output.png --annotate --session s1 --tab t1  # With labels
actionbook browser pdf output.pdf --session s1 --tab t1               # Save as PDF

Interaction

# Click
actionbook browser click "<selector>" --session s1 --tab t1           # Click element
actionbook browser click 420,310 --session s1 --tab t1                # Click coordinates
actionbook browser click @e5 --session s1 --tab t1                    # Click by snapshot ref
actionbook browser click "<selector>" --count 2 --session s1 --tab t1  # Double-click

# Text input
actionbook browser fill "<selector>" "text" --session s1 --tab t1     # Clear field, then set value
actionbook browser type "<selector>" "text" --session s1 --tab t1     # Type keystroke by keystroke

# Keyboard
actionbook browser press Enter --session s1 --tab t1
actionbook browser press Control+A --session s1 --tab t1

# Selection
actionbook browser select "<selector>" "value" --session s1 --tab t1
actionbook browser select "<selector>" "Display Text" --by-text --session s1 --tab t1
actionbook browser select "<selector>" @e12 --by-ref --session s1 --tab t1

# Mouse
actionbook browser hover "<selector>" --session s1 --tab t1
actionbook browser focus "<selector>" --session s1 --tab t1
actionbook browser mouse-move 420,310 --session s1 --tab t1
actionbook browser cursor-position --session s1 --tab t1
actionbook browser drag "<source>" "<target>" --session s1 --tab t1

# Scroll
actionbook browser scroll down --session s1 --tab t1
actionbook browser scroll down 500 --session s1 --tab t1
actionbook browser scroll into-view @e8 --session s1 --tab t1
actionbook browser scroll top --session s1 --tab t1

# JavaScript
actionbook browser eval "document.title" --session s1 --tab t1
actionbook browser eval "document.querySelectorAll('a').length" --session s1 --tab t1
actionbook browser eval "await fetch('/api/data').then(r => r.json())" --no-isolate --session s1 --tab t1
actionbook browser eval --file script.js --session s1 --tab t1
echo 'document.title' | actionbook browser eval - --session s1 --tab t1

# File upload
actionbook browser upload "<selector>" /path/to/file.pdf --session s1 --tab t1

fill clears the field and sets the value directly (like pasting). type simulates individual keystrokes and appends to existing content.

eval isolates let/const scope by default. Use --no-isolate for multi-statement async expressions or when you need shared scope across calls.

eval accepts the expression from three mutually-exclusive sources: a positional argument, --file <path>, or stdin (-). Providing more than one (or none) returns EVAL_ARGS_CONFLICT.

On failure, eval returns a structured error envelope with error.code set to one of:Runtime errors (after browser execution):

EVAL_RUNTIME_ERROR — JS exception (ReferenceError, TypeError, etc.)
EVAL_CROSS_ORIGIN — cross-origin fetch or SecurityError
EVAL_RESPONSE_NOT_JSON — response Content-Type is not JSON
EVAL_RESPONSE_NOT_OK — HTTP status is not 2xx
EVAL_TIMEOUT — expression did not resolve within --timeout

CLI-layer errors (before any browser interaction):

EVAL_ARGS_CONFLICT — multiple input sources, or no source at all
EVAL_FILE_NOT_FOUND — --file path unreadable (not found, permission denied, invalid data)
EVAL_STDIN_TTY — positional - but stdin is a terminal
EVAL_STDIN_EMPTY — stdin read produced empty or whitespace-only input

Each error includes error.hint and error.details with diagnostic fields (reason, and for fetch errors: status, content_type, body_head). Read error.code to branch on the failure class instead of parsing the message string.

Browser commands that interact with elements, navigate, or communicate via CDP return structured CDP_* error codes:

CDP_NODE_NOT_FOUND — DOM node is stale or nonexistent (retryable: no)
CDP_NOT_INTERACTABLE — element exists but can’t be acted on (retryable: no)
CDP_NAV_TIMEOUT — navigation or eval timeout (retryable: yes)
CDP_TARGET_CLOSED — CDP target closed mid-command (retryable: yes)
CDP_PROTOCOL_ERROR — CDP response malformed or missing fields (retryable: no)
CDP_GENERIC — unclassified CDP error (retryable: no)

When error.code is a CDP_* code, error.details includes reason (raw CDP message) and cdp_code (upstream numeric code) when available. CDP_NAV_TIMEOUT also includes timeout_ms. Each code (except CDP_GENERIC) includes an actionable error.hint.Some interaction paths still emit the legacy CDP_ERROR code — these are being migrated to the structured taxonomy (ACT-999).

Wait

actionbook browser wait element "<selector>" --session s1 --tab t1
actionbook browser wait navigation --session s1 --tab t1
actionbook browser wait network-idle --session s1 --tab t1
actionbook browser wait condition "document.readyState === 'complete'" --session s1 --tab t1

Default timeout: 30000ms. Override with --timeout <ms>.

wait network-idle is edge-triggered: it only tracks fetch/XHR requests started after the command begins. Pre-existing background connections (SSE, WebSocket, in-flight fetches) are ignored and do not block. This is an agent-friendly settle signal, not a guarantee of global network silence.

Logs & Network

# Console logs
actionbook browser logs console --session s1 --tab t1
actionbook browser logs console --level warn,error --session s1 --tab t1
actionbook browser logs errors --session s1 --tab t1

# Network
actionbook browser network requests --session s1 --tab t1
actionbook browser network requests --filter /api/ --method POST --session s1 --tab t1
actionbook browser network requests --dump --out /tmp/dump --session s1 --tab t1  # Export to requests.json
actionbook browser network request <id> --session s1 --tab t1         # Full detail + response body

# HAR recording
actionbook browser network har start --session s1 --tab t1            # Start recording
actionbook browser network har stop --session s1 --tab t1             # Stop and export HAR 1.2 file
actionbook browser network har stop --session s1 --tab t1 --out /tmp/trace.har  # Custom output path

HAR recording is per-tab. Multiple tabs or sessions can record independently. har start accepts --max-entries N to set the ring-buffer cap (default: 10000). Output is HAR 1.2 JSON with request/response headers and timings (no response bodies — use --dump for that). If --out is omitted, a timestamped file is created in ~/.actionbook/har/. Redirect chains produce one entry per hop. Error codes: HAR_ALREADY_RECORDING, HAR_NOT_RECORDING.

When har stop completes and entries were dropped due to the ring-buffer cap, the envelope signals truncation:

meta.truncated == true
meta.warnings contains "HAR_TRUNCATED: <N> earlier entries dropped (max_entries=<cap>); raise --max-entries or stop recording sooner to keep the full trace"
data.max_entries — the configured cap at stop time

On a clean stop (dropped == 0), meta.truncated is false and meta.warnings is empty. data.path, data.count, and data.dropped are always present.

Cookies & Storage

# Cookies (session-level, no --tab)
actionbook browser cookies list --session s1
actionbook browser cookies get session_id --session s1
actionbook browser cookies set token abc123 --session s1
actionbook browser cookies delete token --session s1
actionbook browser cookies clear --session s1

# Local Storage
actionbook browser local-storage list --session s1 --tab t1
actionbook browser local-storage get myKey --session s1 --tab t1
actionbook browser local-storage set myKey "value" --session s1 --tab t1

# Session Storage (same syntax as local-storage)
actionbook browser session-storage list --session s1 --tab t1

Batch Operations

actionbook browser batch-new-tab --urls https://a.com https://b.com --session s1
actionbook browser batch-snapshot --tabs t1 t2 t3 --session s1
actionbook browser batch-click @e5 @e6 @e7 --session s1 --tab t1

`actionbook extension`

Manage the Chrome extension used by extension mode. The extension bridge runs inside the actionbook daemon (auto-started by browser commands). The recommended install method is the Chrome Web Store (current version: 0.4.0). actionbook extension install is a local fallback — after running it, you must manually load the unpacked extension in Chrome via chrome://extensions > Developer mode > Load unpacked, using the path from actionbook extension path.

actionbook extension status                   # Bridge status + extension connection state
actionbook extension ping                     # Measure bridge RTT
actionbook extension install                  # Fallback: install to ~/Actionbook/extension/ (requires manual Chrome load)
actionbook extension install --force          # Force reinstall
actionbook extension uninstall                # Remove extension
actionbook extension path                     # Print install path, status, and version

extension status returns the bridge state (listening, not_listening, or failed) and whether the extension is connected. extension ping connects to the bridge WebSocket at ws://127.0.0.1:19222 and measures round-trip time.

Extension 0.4.0 changes: Tabs opened by Actionbook are automatically grouped into a Chrome tab group titled “Actionbook” (toggleable via the extension popup). In extension mode, list-tabs now returns only Actionbook-managed tabs (debugger-attached or in the Actionbook tab group) — other user tabs are hidden. Extension versions below 0.4.0 are rejected at handshake.

`actionbook daemon`

The actionbook daemon runs in the background and manages browser sessions. It auto-starts on the first CLI call.

actionbook daemon restart                     # Stop the running daemon (next CLI call respawns)

​CLI Reference

​actionbook search

​actionbook get

​actionbook manual

​actionbook setup

​actionbook browser

​Global Flags

​Session Lifecycle

​actionbook browser start

​Other Session Commands

​Tab Management

​Navigation

​Observation

​Interaction

​Wait

​Logs & Network

​Cookies & Storage

​Batch Operations

​actionbook extension

​actionbook daemon

CLI Reference

`actionbook search`

`actionbook get`

`actionbook manual`

`actionbook setup`

`actionbook browser`

Global Flags

Session Lifecycle

`actionbook browser start`

Other Session Commands

Tab Management

Navigation

Observation

Interaction

Wait

Logs & Network

Cookies & Storage

Batch Operations

`actionbook extension`

`actionbook daemon`