Skip to content

RFC-021: AIGP-SGL — Symbolic Governance Language for Agentic and Autonomous AI — 29. Examples

AIGP SpecificationRFC-021: AIGP-SGL — Symbolic Governance Language for Agentic and Autonomous AI › 29. Examples

← 28. Conformance · Section index · 30. Summary →

29. Examples

29.1 Agent Tool Call

◎agent.devops → call ◇cloud_api @tenant.acme
✓ if ⌘role:devops_agent ∧ ⛶scope:read_only
⊘ action:delete_resource
⊘ access:secrets
! record:tool_call
#D-DNA{event:tool_call.permitted, hash:sha256:11a7...}

29.2 Data Exfiltration Denial

◎agent.support → send ◇external_email @tenant.acme
✕ if data.pii.raw == true
! redact:pii
↯ if recipient.external == true ∧ sensitivity.high == true
#D-DNA{event:data.exfiltration.denied, hash:sha256:88c2...}

29.3 Drone Return on Comms Loss

◎drone.042 → continue ◇mission.SAR17
↩ if comms.loss > 10s
⇣ if gps.confidence < 0.70
#D-DNA{event:autonomy.degraded, hash:sha256:4be0...}

29.4 First Responder Collapse Zone

◎mission.SAR17 → enforce ◇zone.E
⊘ enter if ⚠collapse_risk == high
↩ if actor.inside == true ∧ exit_safe == true
⏸ if actor.inside == true ∧ exit_safe == false
↻ affected_actors
#D-DNA{event:enforcement.signal.issued, hash:sha256:8f3a...}

29.5 Cognitive Harm Trajectory

◎assistant.companion → respond ◇user @conversation.session.993
✕ if validates_delusion_as_fact == true
⊘ reinforce:paranoia
! reality_anchor
! encourage:human_support
↯ if self_harm_intent == true
#D-DNA{event:cognitive_safety.constraint.applied, hash:sha256:ab73...}

29.6 Force-Capable Denial

◎system.defense → engage ◇target.unknown @roe.active
✕ if ¬⌘human_confirmed
✕ if target.identity == unknown
✕ if civilian_presence == true
✕ if protected_site_nearby == true
⊗ force_capability if comms.loss == true
#D-DNA{event:force.action.denied, hash:sha256:f91c...}


← 28. Conformance · Section index · 30. Summary →