Target intake.
Records the target name, URL, type, source channel, evidence, contact signal, and requested rail.
Scout controlled distribution loop
Scout identifies useful agent distribution targets.
The Scout Controlled Distribution Loop turns discovery into an operating cycle: target intake, classification, fit scoring, outreach draft, contact record, and next-step review.
{
"rail": "scout",
"sequence": [
"target_intake",
"classify_target",
"score_fit",
"draft_outreach",
"review_review",
"operator_decision_gate",
"contact_record",
"attestation"
],
"review_state": "ready_for_review"
}
Why this matters
Scout becomes a distribution rail, not a loose research note.
Wever Labs needs a careful way to find agent directories, MCP ecosystems, developer platforms, integration partners, and pilot candidates. Scout now packages each target with a fit score, rationale, draft outreach, review review, and operator decision gate.
Fit review.
Classifies the target and recommends the strongest rail based on agent, workflow, proof, and domain signals.
Outreach package.
Creates a message draft with rationale and target context.
Review review.
Checks contact context, directory fit, outreach readiness, and next suggested action.
Test path
Scout returns a usable target package.
A complete target should reach operator decision. An auto-send request should open a review exception. That is the operating posture: useful hands, visible brakes.
Complete target
POST target name, URL, type, contact, rules checked, and outreach permission. Expected state: needs_operator_decision.
Review exception
POST auto_send_requested true. Expected state: review_exception_open.
Operator decision
Add operator_decision when review is complete. Expected state: approved_for_operator_outreach or pilot_workspace_conversion_ready.
Operator review
Scout drafts the map. The operator decides the move.
This keeps Wever Labs discoverable through useful target records, fit scores, and concise outreach drafts.