返回 Skills

TinyFish

oo-tinyfish

TinyFish (tinyfish.ai). Use this skill for ANY TinyFish request — reading, creating, updating, and deleting data. Whenever a task involves TinyFish, use this skill instead of calling the API directly.

描述與指示可能以英文或作者原始語言顯示。

SKILL.md

TinyFish

Operate TinyFish through your OOMOL-connected account. This skill calls the tinyfish connector with the oo CLI; OOMOL injects credentials server-side, so you never handle raw tokens.

Running an action

Assume the user has already installed the oo CLI, signed in, and connected TinyFish. Do not run oo auth login or open the connection URL proactively — just run the action. Fall back to First-time setup only when a command actually fails with an auth or connection error.

1. Inspect the contract to get the authoritative input/output schema before building a payload:

oo connector schema "tinyfish" --action "<action_name>"

2. Run the action with a JSON payload that matches the input schema:

oo connector run "tinyfish" --action "<action_name>" --data '<json>' --json
  • --data takes a JSON object string or @path/to/file.json; omit it to send {}.
  • The response is { "data": ..., "meta": { "executionId": "..." } }; the execution id lives under meta.executionId.

Each action is listed below with a one-line description; actions that change state carry a [write] or [destructive] tag. Before constructing --data, fetch the action’s live schema with oo connector schema to get its authoritative input fields.

Available actions

  • cancel_profile_setup_session — Cancel a profile setup session without saving browser state. [destructive]
  • cancel_research_run — Cancel a research run and return whether its state changed. [destructive]
  • cancel_run — Cancel an asynchronous automation run. Already terminal runs retain their status. Runs started through the synchronous API cannot be cancelled. [destructive]
  • cancel_runs — Cancel up to 100 automation runs. Already terminal runs retain their existing status; missing IDs are returned separately. [destructive]
  • connect_vault — Connect 1Password or Bitwarden and immediately sync display-safe credential metadata. [write]
  • create_browser_session — Create a remote browser session and return CDP connection details for an external browser client. [write]
  • create_monitor — Create a Page or Topic Monitor with a schedule and baseline run; requires a personal API key. [write]
  • create_profile — Create a Browser Context Profile for reusable browser state. [write]
  • create_profile_setup_session — Start an interactive setup browser to sign in and capture state for a profile. [write]
  • delete_monitor — Delete a Monitor and stop its scheduled runs. [destructive]
  • delete_profile — Delete a Browser Context Profile and its saved state. [destructive]
  • delete_profile_domain — Delete cookies and storage for one saved profile domain. [destructive]
  • disconnect_vault — Disconnect a password manager and remove its stored enabled items. [destructive]
  • fetch_urls — Fetch up to ten URLs with content extraction, selectors and conditional request support; inspect per-URL errors.
  • get_monitor — Get a Page or Topic Monitor and its configuration.
  • get_profile — Get a Browser Context Profile and its saved domain metadata.
  • get_research_run — Get research progress, final reports, sources and terminal outcome.
  • get_run — Retrieve an automation run and captured steps. Prefer screenshot URLs for JSON clients; use get_step_screenshot or get_step_html to transfer captured files.
  • get_runs — Retrieve up to 100 automation runs, including per-run results, steps, and missing IDs.
  • get_step_html — Download a captured automation-step HTML snapshot and return a file-transit URL. Requires capture_config.html when submitting the run.
  • get_step_screenshot — Download a captured automation-step screenshot and return a file-transit URL.
  • get_wallet — Get wallet balance, auto-reload configuration, rates, and pending top-up details.
  • list_browser_usage — List paginated browser session usage, filtered by session, timestamps, or status.
  • list_fetch_usage — List paginated fetch usage records.
  • list_monitors — List Page and Topic Monitors owned by the personal API key.
  • list_profiles — List Browser Context Profiles and tracked signed-in sites.
  • list_research_runs — List research runs with query, state, timestamp and cursor filters.
  • list_runs — List and search automation runs with cursor pagination, status, goal, and creation-time filters.
  • list_search_usage — List paginated search usage records and recorded results.
  • list_vault_connections — List connected password managers and validation status.
  • list_vault_items — List display-safe vault item metadata available from connected password managers.
  • run_monitor — Run a Monitor immediately and return its Page or Topic results. [write]
  • save_profile_setup_session — Save cookies and browser storage from a setup session into its profile. [write]
  • search — Search web pages, news or research papers with optional content fetching.
  • start_research — Start deep or maximum-depth research and retrieve its reports through get_research_run. [write]
  • submit_run — Submit a browser automation task asynchronously. The agent can perform arbitrary website operations; use get_run to retrieve progress and results. [destructive]
  • submit_run_batch — Atomically submit 1–100 browser automation tasks. Shared profile_id and output_schema apply to the batch. Retrieve each run using get_run or get_runs. [destructive]
  • sync_vault_items — Sync connected password managers and return item metadata and synchronization counts. [write]
  • terminate_browser_session — Terminate a remote browser session; an already-ended session also succeeds. [destructive]
  • top_up_wallet — Request a wallet top-up or submit caller-supplied MPP payment authorization. Return the payment challenge when payment is required. A pending result means payment was charged; do not pay again. [write]
  • update_monitor — Pause or resume a Monitor, or update its schedule and optional metadata; these update modes cannot be combined. [destructive]
  • update_profile — Update profile metadata, default selection, or proxy country. [write]
  • upload_profile_state — Upload cookies, local storage, and session storage into a Browser Context Profile. [write]

Safety

  • Untagged actions are reads (get / list / search) — safe to run directly.
  • Actions tagged [write] change TinyFish state — confirm the exact payload and effect with the user before running.
  • Actions tagged [destructive] remove or overwrite data — always confirm the target and get explicit approval first.

First-time setup

These are one-time steps — do not repeat them on every call. Run a step only when a command fails for the matching reason.

  • oo: command not found — install the oo CLI (other platforms: https://cli.oomol.com/install-guide.md):

    curl -fsSL https://cli.oomol.com/install.sh | bash    # macOS / Linux
    
    irm https://cli.oomol.com/install.ps1 | iex           # Windows PowerShell
    
  • Not signed in / authentication error — sign in to your OOMOL account once:

    oo auth login
    
  • scope_missing / credential_expired / app_not_ready / app_not_found — TinyFish is not connected, or the connection expired or lacks a scope. Connect once (auth type: API key) at:

    https://console.oomol.com/app-connections?provider=tinyfish
    
  • HTTP 402 / OOMOL_INSUFFICIENT_CREDIT — billing stop. Recharge at https://console.oomol.com/billing/token-recharge before retrying.

Resources

想在 Slack 或 Teams 中使用 TinyFish?

TinyFish

一鍵在 Leina 中運行