Arc Skills / Technical management and design reviews
Perform a technical risk assessment
Use Arc Skills to identify credible technical risks from a design, its interfaces and its evidence gaps. The output explains the causal path to a project objective, separates existing issues from future events and proposes controls without inventing scores.
Use this skill
Use Arc Skills to perform a technical risk assessment.
Inputs:
- Project objectives, current design and configuration
- Interface assumptions, supplier evidence and open anomalies
- Existing controls and any approved risk scales
Return cause → uncertain event → consequence scenarios with evidence, trigger, control weakness, proposed response and owner. Separate existing issues from risks. Do not assign scores unless the project’s scale is provided.What you provide and what you get
| What you have | How it is used | What you get |
|---|---|---|
| Objectives and design evidence | Identify credible departures | Specific causal risk scenarios |
| Open assumptions and supplier dependencies | Check uncertain interfaces and maturity | Evidence-backed omissions and triggers |
| Controls and approved scales | Assess response options | Unscored or correctly scored actions |
A sensor integration risk comes from a specific gap
Illustrative engineering example.
The design needs a sensor output within an interface noise limit. The supplier has tested a laboratory unit, while the flight harness and switching regulator have not been tested together.
Objective: acquire sensor data within the controlled accuracy requirement.
Evidence: sensor report REP-8 uses short laboratory wiring; flight harness H3 is longer and runs beside a switching regulator.
Existing issue: connector drawing INT-4 has a missing shield termination definition.
No project likelihood or consequence scale supplied.| Scenario | Evidence / cause | Uncertain event and consequence | Trigger / control | Proposed response and owner |
|---|---|---|---|---|
| Harness coupling | REP-8 omits H3 and regulator environment | Coupled noise may exceed the sensor error allocation, degrading measurement accuracy | Trigger: representative harness noise result crosses the approved limit | Electrical owner tests the actual routing and assesses filtering or separation |
| Late interface correction | Shield termination undefined in INT-4 | After integration, a required grounding change may force harness rework | Resolve drawing omission before harness release | Interface owner completes the controlled termination agreement |
| Supplier configuration change | Flight unit configuration not compared with laboratory unit | Unreviewed differences may invalidate assumed sensor performance | Trigger: delivered configuration differs from tested unit | Supplier owner obtains version comparison and applicability evidence |
The missing shield definition is already an issue. The future rework it could cause is a risk scenario with a different action and trigger. Keeping both visible avoids pretending that documenting a risk fixes the current drawing.
No probability or severity number is supportable from the supplied information. The assessment remains useful because each scenario has a specific evidence gap and an action that can reduce uncertainty or interrupt the causal path.
Build scenarios from evidence, not a generic risk list
- Name the objective and configuration whose performance could be affected.
- Write cause, future uncertain event and consequence; separate present issues.
- Check existing controls, triggers and cross-interface or supplier dependencies.
- Apply only supplied risk scales and propose a response with an accountable owner.
Questions about this task
Is missing evidence always a technical risk?
No. Explain the plausible event and consequence. A missing document without a technical or project mechanism is an information gap, not automatically a high risk.
Sources and further reading
- NASA Systems Engineering Handbook: technical risk management: Background on risk scenarios, uncertainty, consequences and responses.
References inspected on 3 October 2026. The worked output is an authored example using the stated inputs.