schema_version: "1.1"
id: adtech/chatgpt-ads-experiment-supervisor
name: ChatGPT Ads Experiment Supervisor
description: Coordinate task-led ChatGPT advertising experiments across creative, policy, measurement ownership, platform inspection, reconciliation, and memory.
version: 0.2.0
released_at: "2026-09-26T00:00:00Z"
category: adtech-agents/experimentation
tags:
  - chatgpt-ads
  - experiment-supervisor
  - measurement
  - agentic-commerce
  - governance
license: MIT
author:
  name: AI Skills Guide Maintainers
  url: https://github.com/ai-knowledge-hub/ai-skills-guide
repository: https://github.com/ai-knowledge-hub/ai-skills-guide
runtimes:
  - codex
  - claude
  - generic
entrypoints:
  spec: AGENT.md
  readme: README.md
  tests: tests/test-prompts.md
  examples_dir: examples
dependencies:
  agents:
    - marketing/creative-operating-system-supervisor
    - adtech/ad-platform-control-plane-supervisor
  skills:
    - marketing/task-map-designer
    - marketing/ai-output-eval-scorecard
    - adtech/advertising-signal-ownership-auditor
    - marketing/agent-share-of-choice-evaluator
    - agentops/ad-platform-policy-gate-designer
  tools:
    - adtech/conversion-event-reconciler
    - adtech/openai-ads-api-client
    - adtech/ad-platform-executor-template
orchestration:
  model:
    selection: runtime-selected
    required_capabilities: [tool-use, structured-output]
  memory:
    mode: bundled
    path: config/memory-profile.default.json
    required: true
  governance:
    mode: bundled
    path: config/governance.default.json
    required: true
  bindings:
    - {name: creative_supervisor, kind: agent, package: marketing/creative-operating-system-supervisor, version: 0.3.0, requirement: required, availability: resolved}
    - {name: control_plane_supervisor, kind: agent, package: adtech/ad-platform-control-plane-supervisor, version: 0.2.0, requirement: required, availability: resolved}
    - {name: task_map_design, kind: skill, package: marketing/task-map-designer, version: 0.1.0, requirement: required, availability: resolved}
    - {name: output_evaluation, kind: skill, package: marketing/ai-output-eval-scorecard, version: 0.1.0, requirement: required, availability: resolved}
    - {name: signal_ownership, kind: skill, package: adtech/advertising-signal-ownership-auditor, version: 0.1.0, requirement: required, availability: resolved}
    - {name: share_of_choice, kind: skill, package: marketing/agent-share-of-choice-evaluator, version: 0.1.0, requirement: required, availability: resolved}
    - {name: policy_gate_design, kind: skill, package: agentops/ad-platform-policy-gate-designer, version: 0.1.0, requirement: required, availability: resolved}
    - {name: reconcile_conversion_events, kind: tool, package: adtech/conversion-event-reconciler, version: 0.1.0, requirement: required, access: read-only, availability: resolved}
    - {name: inspect_openai_ads, kind: tool, package: adtech/openai-ads-api-client, version: 0.2.0, requirement: required, access: read-only, availability: resolved}
    - {name: execute_ad_platform_plan, kind: tool, package: adtech/ad-platform-executor-template, version: 0.2.0, requirement: optional, access: read-write, availability: resolved}
operational:
  role: Experiment supervisor for task-led conversational advertising and agent-mediated commerce.
  coordinates:
    - Task hypothesis design
    - Creative and compliance readiness
    - Signal ownership and measurement
    - Mock or read-only platform inspection
    - Event reconciliation and experiment memory
  autonomy_level: advisory
  approval_boundary: May plan, inspect approved read-only data, and reconcile supplied evidence; live campaign writes require the existing policy gate, executor, and human approval.
  outputs:
    - Experiment plan
    - Readiness decision
    - Reconciled evidence
    - Experiment memory record
usability:
  availability: setup-required
  execution: orchestrator
  limitations:
    - Provider authentication and the pinned control-plane dependencies must pass readiness before connected use.
    - The supervisor is advisory and does not execute campaign writes.
  quickstart: Install the agent and begin with its mock workflow; use the compiled contract to configure connected providers.
verification:
  tests_min_prompts: 5
  security_reviewed: false
deprecated: false