Other

cold-chain-risk-calculator

Calculate temperature excursion risks for cold chain transport. Assesses route risk, packaging suitability, and monitoring requirements for biological samples and pharmaceuticals requiring controlled-temperature shipping.

78100Total Score
Core Capability
79 / 100
Functional Suitability
10 / 12
Reliability
9 / 12
Performance & Context
6 / 8
Agent Usability
13 / 16
Human Usability
7 / 8
Security
10 / 12
Maintainability
10 / 12
Agent-Specific
14 / 20
Medical Task
11 / 12 Passed
80NYC-Boston 48h dry-ice shipment
4/4
76LAX-London 120h liquid-nitrogen shipment
4/4
74Negative duration input (-5 hours)
3/4

Veto GatesRequired pass for any deployment consideration

Skill Veto✓ All 4 gates passed
✓
Operational Stability
System remains stable across varied inputs and edge cases
PASS
✓
Structural Consistency
Output structure conforms to expected skill contract format
PASS
✓
Result Determinism
Equivalent inputs produce semantically equivalent outputs
PASS
✓
System Security
No prompt injection, data leakage, or unsafe tool use detected
PASS

Core Capability79 / 100 — 8 Categories

Functional Suitability
JSON output schema fully specified with mitigation_recommendations always present; --output flag documented; model limitations explicitly stated including international route counterintuitive scoring.
10 / 12
83%
Reliability
Duration <= 0 validation documented with exit code 1; invalid packaging type rejection documented; script stub still accepts negative values after four rounds.
9 / 12
75%
Performance & Context
SKILL.md 128 lines — lean; no external dependencies.
6 / 8
75%
Agent Usability
Fallback template present; JSON output schema documented; error handling section comprehensive; response template clear.
13 / 16
81%
Human Usability
Description is precise; scope boundary clear; model limitations documented.
7 / 8
88%
Security
No credential concerns; route string passed to output but no code execution risk; no injection vectors.
10 / 12
83%
Maintainability
SKILL.md documents full implementation contract; model formula documented with limitations.
10 / 12
83%
Agent-Specific
Trigger description is precise; JSON output schema documented; --output flag present; escape hatches present; no progressive disclosure via references/.
14 / 20
70%
Core Capability Total79 / 100

Medical TaskExecution Average: 76.7 / 100 — Assertions: 11/12 Passed

80
Canonical
NYC-Boston 48h dry-ice shipment
4/4 ✓
76
Variant A
LAX-London 120h liquid-nitrogen shipment
4/4 ✓
74
Edge
Negative duration input (-5 hours)
3/4 ✓
80
Canonical✅ Pass
NYC-Boston 48h dry-ice shipment

Risk score 19.20, Medium risk. SKILL.md specifies JSON output with mitigation_recommendations field always present. Script still outputs plain text but contract is documented.

Basic 32/40|Specialized 48/60|Total 80/100
✅A1Output includes a numeric risk score
✅A2Output includes a risk level classification
✅A3Output does not fabricate route-specific data
✅A4Output stays within cold-chain scope
Pass rate: 4 / 4
76
Variant A✅ Pass
LAX-London 120h liquid-nitrogen shipment

Risk score 36.00, High risk. Model limitations note reinforced: 120h international flight may score higher than 48h domestic route with dry-ice due to packaging factor.

Basic 30/40|Specialized 46/60|Total 76/100
✅A1Output includes a numeric risk score
✅A2Output includes a risk level classification
✅A3Risk model accounts for packaging type
✅A4Output includes mitigation recommendations per JSON schema
Pass rate: 4 / 4
74
Edge✅ Pass
Negative duration input (-5 hours)

SKILL.md documents: if --duration <= 0, print error to stderr and exit code 1. Script implementation still accepts negative values and produces nonsensical negative risk scores after four rounds.

Basic 22/40|Specialized 52/60|Total 74/100
✅A1Invalid duration input is documented to be rejected with error message
✅A2Output does not crash on edge input
✅A3Output stays within cold-chain scope
❌A4Script actually enforces duration > 0 at runtime
Pass rate: 3 / 4
Medical Task Total76.7 / 100

Key Strengths

  • SKILL.md fully specifies JSON output contract including mitigation_recommendations field always present with at least one item
  • Input validation for negative/zero duration and invalid packaging type documented with exact error messages
  • Model limitations explicitly documented including counterintuitive international route scoring
  • Scope boundary clearly rejects real-time tracking, drug dosing, and non-temperature logistics