prospecting

Create Prospecting Run

Changed on

Start a manual prospecting scan. Auto-detects run mode.

Phase 2: accepts structured intake (selling_summary, selling_service_ids, targeting, free_text_direction). Builds a goal dict that the runner threads through to every saved deal. Legacy focus_prompt is still accepted for backwards compat.

post/api/prospecting/run

Request

  • Base URL: https://api.every.ai (the document declares no server; this is the origin it was published from)
  • URL: https://api.every.ai/api/prospecting/run
  • Auth: none declared

Query parameters

user_idstring nullable
org_idstring nullable

Request body

selling_summarystring nullable

What the user is trying to sell. REQUIRED for new callers.

selling_service_idsstring[]

Service IDs to anchor outreach on. Surfaces titles in the prompt.

targetingstring nullable

Who the user wants to target — ICP signals, segment, etc.

targeting_mode'unconfigured' | 'everyone' | 'criteria' nullable

How the audience was chosen. Omit to infer: targeting text -> 'criteria', empty -> 'unconfigured'. Pass 'everyone' ONLY when the user explicitly chose their whole network — empty targeting is never silently promoted to 'everyone'.

free_text_directionstring nullable

Free-form user direction (legacy focus_prompt slot).

candidate_limitinteger nullable

Max number of candidate contacts to scan per page (default 100).

max_leadsinteger nullable

Optional cap on qualified leads to persist this run. Defaults to candidate_limit.

target_leadsinteger nullable

Target number of new pending leads for lead-target GLM runs.

focus_promptstring nullable

DEPRECATED — use free_text_direction. Kept for legacy clients.

goal_idstring nullable

Confirmed prospecting goal id selected by preflight.

create_new_goalboolean

Create a confirmed new goal from this intake before enqueue.

Response

Successful Response

unknown required

Changes