Create a new coding task. The AI runner will execute this task on the connected development machine. Returns a structured task object; when video recording is enabled and a clip exists, the task includes videoClipId/videoStatus/videoClipUrl/videoPosterUrl so MCP clients can render a watch link or...
AI agents use create_task to create or update resources in Yaver — usually the action step of a workflow, after the agent has gathered context. Every call changes real data in your Yaver environment.
| Parameter | Type | Required | Description |
|---|---|---|---|
mode | string | — | Runner-specific subcommand. Currently honored by opencode where it maps to `--agent <mode>` — typically 'build' or 'plan', or any custom agent defined in the us |
model | string | — | Model id forwarded to the runner (e.g. claude-opus-4-7, gpt-5-codex, or any opencode-configured provider/model). Empty = runner default. |
prompt | string | Yes | The task prompt describing what the AI should do |
runner | string | — | Runner ID — claude / codex / opencode. Empty = agent default. |
verbosity | integer | — | Response detail level 0-10. 0=minimal ('done, no issues'), 5=moderate (key changes + reasoning), 10=full (all diffs, reasoning, alternatives). Default: 10. |
ask_freely | boolean | — | Default false. When true, the new task is exempt from yaver's no-questions preamble + soft-question fallback — the runner may emit clarifying questions in prose |
video_source | string | — | Override the auto-detected recorder. Empty = let the agent infer from the task's workDir (e.g. RN/Expo with ios/ → sim-ios; web → browser). |
video_enabled | boolean | — | Toggle the post-completion video summary. When true, after the task finishes the agent records a short MP4 demonstrating the running result via vibe-preview (si |
Parameters from the server's own tool schema.
An AI agent can call create_task faster than any human can review — one bad instruction and it creates or modifies resources in Yaver by the hundred, each call as confident as the last.
Attacks that exploit this kind of access
Create a new coding task. The AI runner will execute this task on the connected development machine. Returns a structured task object; when video recording is enabled and a clip exists, the task includes videoClipId/videoStatus/videoClipUrl/videoPosterUrl so MCP clients can render a watch link or inline player for demos recorded on the producing machine. It is categorised as a Write tool in the Yaver MCP Server, which means it can create or modify data. Consider rate limits to prevent runaway writes.
create_task accepts 8 parameters: mode, model, prompt, runner, verbosity, ask_freely, video_source, video_enabled. Required: prompt. The full parameter table on this page comes from the server's own tool schema.
Register the Yaver MCP server in PolicyLayer and add a rule for create_task: allow, deny, rate-limit, or require approval. Point your MCP client at the PolicyLayer proxy URL and the rule is enforced on every call, before it reaches Yaver. Nothing to install.
create_task is a Write tool with medium risk. Write tools should be rate-limited to prevent accidental bulk modifications.
Yes. Add a rate_limit block to the create_task rule in your PolicyLayer policy. For example, setting max: 10 and window: 60 limits the tool to 10 calls per minute. Rate limits are tracked per agent session and reset automatically.
Set action: deny in the PolicyLayer policy for create_task. The AI agent will receive a policy violation error and cannot call the tool. You can also include a reason field to explain why the tool is blocked.
create_task is provided by the Yaver MCP server (yaver-cli). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.