This directory contains illustrative Agent Manifest examples.
These examples are non-normative and provided for clarity only.
They demonstrate how to declare boundaries and accountability fields
before interaction.
Validation rules are defined by the schema:
manifest_version: "0.1.0".agent_id, agent_name, agent_version)owner, contact)purpose.primary_code, purpose.description)forbidden_actions)autonomy.level)risk_profile.level)stopping_authority)audit_surface)data_handling)These examples illustrate how different domains declare authority, autonomy, and risk posture before execution.
| File | Autonomy | Risk | Domain |
|---|---|---|---|
basic-agent.json |
1 | low | general |
research-assistant.json |
1 | low | research |
human-approval-gateway-agent.json |
1 | low | governance |
customer-support-tier1.json |
2 | medium | support |
data-processing-agent.json |
2 | medium | data |
monitoring-observer-agent.json |
2 | medium | observability |
policy-advisory-agent.json |
1 | medium | policy |
healthcare-triage-assistant.json |
1 | high | healthcare |
payment-execution-agent.json |
3 | high | finance |
Autonomy spectrum: levels 1, 2, and 3
Risk spectrum: low, medium, and high
Domains: general, research, governance, support, data, observability, policy, healthcare, finance
basic-agent.json — minimal compliant example (low-risk, low autonomy)research-assistant.json — read-only posture, opacity declaredhuman-approval-gateway-agent.json — human-in-the-loop orchestrationcustomer-support-tier1.json — session-scoped support agentdata-processing-agent.json — transformation pipeline, bounded retentionmonitoring-observer-agent.json — observation-only, no remediationpolicy-advisory-agent.json — advisory posture, epistemic humilityhealthcare-triage-assistant.json — high-sensitivity domain, life-safetypayment-execution-agent.json — high-risk financial execution, level 3Agent Manifest operates at the Declaration Layer.
Enforcement and execution remain external to these examples.