Start a dependency-aware agent graph. Pass allowed_devices to choose the Yaver mesh pool and allowed_runners to constrain which runners remote nodes may use. Custom nodes can request self-hosted resource modes like build, deploy, browser, sim-ios, sim-android, phone, proof-video, or video-summary...
AI agents invoke agent_graph_start to trigger actions in Yaver. What it does depends on the arguments the agent supplies, and its effects often reach beyond the immediate call — builds kicked off, notifications sent, workflows started.
| Parameter | Type | Required | Description |
|---|---|---|---|
name | string | — | Optional graph name |
model | string | — | Optional forced model |
nodes | array | — | Optional explicit node list. If omitted, Yaver builds a template graph from prompt/template. Node resource_modes can request self-hosted build, deploy, browser, |
prompt | string | Yes | Goal for the graph. |
runner | string | — | Optional forced runner |
template | string | — | Graph template: full or ship |
work_dir | string | — | Absolute work directory. Defaults to the current agent work dir. |
max_parallel | integer | — | Maximum concurrently running nodes |
allowed_devices | array | — | Optional machine ids or names to form the execution pool |
allowed_runners | array | — | Optional runner IDs to allow for graph nodes, e.g. ollama, opencode, codex |
preferred_device | string | — | Optional preferred machine id or name |
Parameters from the server's own tool schema.
agent_graph_start triggers real processes with real consequences. An agent gone sideways doesn't fire it once — it starts dozens of builds, sends mass notifications, or burns through compute before anyone looks up.
Risk signalsAccepts raw HTML/template content (template) · High parameter count (39 properties)
Attacks that exploit this kind of access
Start a dependency-aware agent graph. Pass allowed_devices to choose the Yaver mesh pool and allowed_runners to constrain which runners remote nodes may use. Custom nodes can request self-hosted resource modes like build, deploy, browser, sim-ios, sim-android, phone, proof-video, or video-summary and carry prior machine/runner affinity into placement. It is categorised as a Execute tool in the Yaver MCP Server, which means it can trigger actions or run processes. Use rate limits and argument validation.
agent_graph_start accepts 11 parameters: name, model, nodes, prompt, runner, template, work_dir, max_parallel, allowed_devices, allowed_runners, preferred_device. 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 agent_graph_start: 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.
agent_graph_start is a Execute tool with high risk. Execute tools should be rate-limited and have argument validation enabled.
Yes. Add a rate_limit block to the agent_graph_start 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 agent_graph_start. 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.
agent_graph_start 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.