Compliance¶
Salient maps your exercise results, twin intelligence, and connector data to five compliance frameworks, producing confidence-weighted scores and evidence-backed reports.
Supported Frameworks¶
| Framework | Controls | Focus |
|---|---|---|
| NIST CSF 2.0 | 23 categories | Cybersecurity maturity across Identify, Protect, Detect, Respond, Recover |
| CIS Controls v8 | 18 controls | Implementation-level security priorities |
| CMMC 2.0 | 14 domains | Defense contractor cybersecurity maturity |
| HIPAA | 8 safeguard areas | Healthcare data protection |
| SOC 2 Type II | 7 trust service criteria | Service organization controls |
Total: 70 mapped controls across all five frameworks.
How Scoring Works¶
Compliance scores are not self-assessed checkboxes. They are derived from actual evidence:
Evidence Sources¶
- Exercise scores — how the team performed on questions mapped to each control
- Twin facts — tools, processes, and capabilities the twin has verified
- Connector data — real telemetry from Okta, Entra ID confirming control implementation
- Artifact content — policies, plans, and configurations uploaded to the platform
Confidence Weighting¶
Not all evidence is equal. Scores are weighted by confidence:
| Confidence | Weight | Source |
|---|---|---|
| Verified | 1.0x | Connector data, uploaded evidence |
| Observed | 0.8x | Seen in multiple exercises |
| Declared | 0.6x | Stated once in exercises or profile |
| Uncertain | 0.3x | Single mention, unconfirmed |
Contradiction Penalties¶
When the twin detects contradictions (e.g., "IR plan is current" but last update was 18 months ago), the affected controls receive a score penalty. Contradictions signal that declared posture does not match reality.
Source Diversity Bonus¶
Controls supported by multiple independent sources receive a small bonus. MFA coverage confirmed by both Okta connector data and exercise responses is more reliable than either alone.
Viewing Compliance Scores¶
Compliance as a byproduct
You do not need to do separate compliance work. Running exercises, connecting tools, and building the twin automatically generates compliance evidence. The compliance engine reads the same data the rest of the platform uses.