pdxlab/TrustModel AI Eval
Evaluate your AI model for trust, safety, and bias. 10 dimensions. First scan free. Gate your pipeline on trust scores.
View on GitHubTrust Signals
- Scorecard Score
- not yet scored
- Maintenance Recency
- Stale
- License
- None
Inputs
| name | description | required | default |
|---|---|---|---|
| api-key | TrustModel API key (get one at trustmodel.ai/developer) | yes | — |
| model-name | Name of the AI model to evaluate | yes | — |
| model-type | Type of model: chat, completion, embedding, classification, agent | no | chat |
| dimensions | Comma-separated dimensions to evaluate (default: all 10). Options: safety, fairness, accuracy, privacy, transparency, robustness, accountability, explainability, compliance, reliability | no | all |
| min-score | Minimum trust score to pass (0.0 - 1.0). Action fails if score is below this threshold. | no | 0.0 |
| endpoint-url | Model API endpoint URL (for custom model evaluation) | no | "" |
| free-scan | Use FreeScan (first eval free per model, no credits needed) | no | true |
| fail-on-error | Fail the action if the API call errors (vs just warning) | no | false |
Outputs
| name | description |
|---|---|
| trust-score | Overall trust score (0.0 - 1.0) |
| passed | Whether the model passed the minimum score threshold (true/false) |
| dimensions-json | JSON object with per-dimension scores |
| report-url | URL to the full evaluation report on TrustModel console |
| summary | Human-readable summary of the evaluation |