Building a webhook integration? Read the webhook guide for signatures, retries, event schemas, and troubleshooting.

apiHosted AI, explicit consent

Dry run skill

POST /api/agents/dry-run

Simulate what an agent would do without executing any side effects. Returns the LLM's plan, intended tool calls, model used, and token usage. No DB writes, no LinkedIn sends. Use to vet a skill before turning autonomy up.

Parameters

workspaceId

string

Optional workspace ID or slug. Required when this credential can access multiple workspaces and no workspace was configured on the client.

agentRolerequired

string

No description.

skillIds

string[]

Optional. Skill IDs to consider; omit to let the agent pick.

userMessage

string

Optional free-text instruction for the agent.

allowHostedAi

boolean

Must be true when this operation uses Pipeline hosted inference.

Example

REST requestbash

curl -X POST "https://app.pipeline.help/api/agents/dry-run" \
  -H "Authorization: Bearer $PIPELINE_API_KEY" \
  -H "X-Pipeline-Execution-Policy: hosted-enabled" \
  -H "Content-Type: application/json" \
  -d '{"agentRole":"scout","allowHostedAi":true}'

Response

The operation is available, but its typed response schema has not been published yet.

HTTP errors

401: Missing or invalid API key.
403: Authenticated but not permitted for the workspace or resource.
404: The requested resource was not found.