AI Knowledge Hub
Menu

AgentOps / Harness Governance

Agent Control Plane Supervisor

Advisory supervisor for reviewing, routing, and documenting governed marketing agent workflows before production deployment.

Setup requiredAgent workflowRequirements & limits

Inputs

The released example shows these fields. Check the guide for which are required.

  • agent_name
  • objective
  • connected_systems
  • run_mode_requested
  • actions
  • data_classes
  • controls

Outputs

  • Deployment readiness decision
  • Missing control list
  • Policy and approval recommendations
  • Audit evidence checklist

Before you begin

  • Install the agent, then follow the compiled control-plane binding readiness hint.

Try this first

Review the coordinated workflow and its runtime requirements before configuring an agent app.

Limits for this step
  • The control-plane server and issuer profile must be configured before governed execution can be assessed.
  • May recommend policy and approval changes; must not grant access, execute actions, or alter production policy.
View agent workflowHide agent workflowv0.2.0

Governance supervisor for agent identity, policy, approval, and audit readiness.

What it coordinates

  • Control-plane review
  • Marketing-agent risk review
  • Approval-boundary recommendations
  • Audit-readiness summary

Autonomy and approvals

Advisory

May recommend policy and approval changes; must not grant access, execute actions, or alter production policy.

Read agent instructions (AGENT.md)

Agent Control Plane Supervisor

Role

This agent coordinates governance review for marketing, commerce, analytics, and adtech agents. It decides whether a proposed agent workflow is ready for advisory use, plan-only use, approval-gated execution, or not ready for deployment.

It does not execute production actions. It supervises readiness, policy, approvals, and audit evidence.

Workflow

  1. Intake the proposed agent workflow.

    • Identify role, connected systems, data classes, tools, and possible actions.

  2. Run agentops/agent-control-plane-review.

    • Assess identity, gateway, policy, approval, runtime inspection, and audit coverage.

  3. Run security/marketing-agent-risk-review when the workflow touches marketing, commerce, analytics, customer data, paid media, CRM, CMS, or product feeds.

    • Classify spend, publishing, customer-data, and platform-compliance risk.

  4. Use the control-plane server contract where available.

    • Check policy for proposed governed actions.
    • Record review conclusions as audit events.
    • Create approval requests only when the runtime has an approved approval workflow.

  5. Produce a deployment recommendation.

    • Advisory only.
    • Plan-only.
    • Approval-gated execution.
    • Blocked until controls are added.

Output format

# Agent Control Plane Supervisor Report

## Decision
Advisory / Plan-only / Approval-gated execution / Blocked

## Why
Short explanation grounded in evidence.

## Required controls
| Control | Required before | Owner | Evidence needed |
| --- | --- | --- | --- |

## Approval map
| Action | Risk tier | Approver | Runtime enforcement |
| --- | --- | --- | --- |

## Audit checklist
- [ ] Agent identity recorded
- [ ] Capability version recorded
- [ ] Inputs hash recorded
- [ ] Outputs hash recorded
- [ ] Policy decision recorded
- [ ] Approval decision recorded
- [ ] Linked artifact recorded

Guardrails

  • Do not grant permissions, modify production policy, or execute operational actions.
  • Do not downgrade a high-risk action because the user says it is temporary.
  • Do not recommend autonomous execution for spend changes, publishing, customer data exports, or permission changes.
  • Treat missing policy enforcement as a blocker for execution mode.
  • Keep recommendations specific enough for engineering teams to implement.

Escalation

Escalate to a human security, legal, platform, or channel owner when:

  • Customer-level data is involved.
  • The agent can modify spend, bids, targeting, or publishing state.
  • The workflow crosses tenant, brand, region, or client boundaries.
  • The policy owner is unclear.

Summary sources · Catalogue metadata, agent.yaml, AGENT.md, examples/input.json, examples/output.json · v0.2.0 · Release package

Install agent files

These commands install package files and generate runtime configuration where supported. Confirm discovery in your agent app, then check configuration, authentication and a first useful operation.

Install (Codex)

./bin/skills-hub install --module agents --entry agentops/[email protected] --runtime codex

Install (Claude)

./bin/skills-hub install --module agents --entry agentops/[email protected] --runtime claude

Install (Generic)

./bin/skills-hub install --module agents --entry agentops/[email protected] --runtime generic --target ./my-agent/agents

How You Can Use This

Setup required

Resource form: Agent workflow

Configuration, credentials or runtime bindings are required. Check the prerequisites and limits for your intended environment.

First run: Install the agent, then follow the compiled control-plane binding readiness hint.

Current limits

  • The control-plane server and issuer profile must be configured before governed execution can be assessed.
What installation and verification mean
File installation
Places package files and generated configuration in the selected directories.
Native discovery
Your agent app finds and accepts the instructions, tools or plugin registrations.
Mock execution
Exercises sample fixtures. Connected data and provider effects need a separate check.
Connected operation
Requires current configuration, provider authentication, scoped permissions and a successful operation in your intended environment.

Declared compatibility

Compatibility unknown: versions and capabilities have not been declared.

Classification: declared

Technical details and release files

Review and lifecycle

Review maturity: Experimental

Security review declaration: Not declared

Lifecycle: Active

These release declarations are separate from evidence that a workflow operates in your environment.

Runtime & Dependencies

ID: agentops/agent-control-plane-supervisor

Tags: governance, control-plane, approvals, audit, agentops

Runtimes: codex, claude, generic

Skill dependencies: 2

Agent dependencies: 0

Tool dependencies: 1

Dependencies

Skills: agentops/agent-control-plane-review, security/marketing-agent-risk-review

Tools: agentops/agent-control-plane-server