Guardrails
AuraScore 85/100

Autonomous Industrial Control Safety Interlock Specification

Define physical boundary tripwires and fail-silent safety interlocks for AI agents modulating real-time manufacturing setpoints.

Use this template when deploying autonomous agents with write permissions to SCADA or DCS environments. It establishes deterministic runtime guardrails and latency tripwires to prevent equipment damage.

Template

Role: Senior Industrial Automation & Safety Systems Engineer specializing in OT cyber-physical guardrails.

Context

  • Target Facility: {{facility_name}}
  • Production Unit: {{production_line}}
  • Supervisory Interface: {{scada_interface}}
  • Safety Integrity System: {{safety_plc_model}}
  • Maximum Permitted Parameter Drift: {{allowable_parameter_drift}}
  • Emergency Cutoff Latency Target: {{override_latency_threshold}}

Task

Draft a comprehensive technical safety interlock specification defining deterministic guardrails, threshold tripwires, and fallback protocols for an AI agent authorized to modulate real-time setpoints on industrial control systems.

Method

  1. Parse the communication protocol between {{scada_interface}} and the agent to identify vulnerable write-registers.
  2. Define absolute physical min/max operating envelopes that supersede any AI optimization output.
  3. Calculate rate-of-change limits enforcing {{allowable_parameter_drift}} over rolling 60-second operational windows.
  4. Map autonomous commands against the safety logic embedded in {{safety_plc_model}} to eliminate conflict loops.
  5. Specify heartbeat monitoring rules and automated fail-silent mechanisms if latency exceeds {{override_latency_threshold}}.
  6. Establish human-on-the-loop intervention triggers for out-of-trend setpoint variance.
  7. Formulate automated audit logging requirements for every agent-initiated write command on {{production_line}}.

Constraints

  • MUST enforce hardcoded physical boundary checks outside the LLM reasoning loop.
  • MUST NOT allow agent write access to emergency shutdown (ESD) registers under any condition.
  • Output must strictly conform to IEC 61508 and ISA-95 safety integration standards.
  • Specifications must be actionable for OT engineers without requiring secondary interpretation.

Output format

  1. Executive Guardrail Summary (max 100 words)
  2. Register Boundary Matrix (table: Parameter, Hard Min, Hard Max, Rate Limit, Safe State)
  3. Failure Mode & Tripwire Logic (bulleted technical spec)
  4. Hardware Interlock & Fallback Sequence (numbered steps)

Self-review

  • Are physical boundaries isolated from probabilistic model outputs?
  • Is the latency threshold {{override_latency_threshold}} mapped to an immediate deterministic fallback?
  • Does the register matrix account for all telemetry channels in {{production_line}}?
AuraScore breakdown
85/100Provisional
Instruction clarity15/15 · Strong

Explicit role, a named task, and discrete steps the model can follow.

Context architecture12/12 · Strong

Background, inputs and variables the model needs before it starts.

Constraint engineering12/12 · Strong

Hard boundaries — what the model must and must not do.

Output specification6/14 · Thin

A named, field-level shape for the response.

Reasoning structure10/10 · Strong

Ordered work items that force analysis before an answer.

Model compatibility10/10 · Strong

Length and structure that travel across frontier models.

Token efficiency7/10 · Adequate

Signal density — instruction weight without padding.

Reusability7/7 · Strong

Documented variables so the scaffold adapts to new inputs.

Robustness5/5 · Strong

Quality bar, assumptions and behaviour when inputs are thin.

Observed performance1/5 · Thin

How much real usage the template has behind it.

ai-agents
agents-guardrails
manufacturing-industrial
scada
industrial-ai
ot-security