> ## Documentation Index
> Fetch the complete documentation index at: https://docs.meshai.dev/llms.txt
> Use this file to discover all available pages before exploring further.
# Readiness Score
> 7-component compliance score (0–120) measuring EU AI Act readiness
The MeshAI readiness score measures your organization's EU AI Act compliance posture across 7 components. The maximum score is **120 points**.
## Score Components
| # | Component | Article | Max Points | What It Checks |
| - | -------------------------- | ---------- | ---------- | ------------------------------------------------------------------------------------------------------------------------------------ |
| 1 | **Audit Trail Active** | Article 12 | 20 | At least one audit event has been recorded for the tenant |
| 2 | **Risk Classification** | Article 6 | 25 | Proportion of registered agents that have a risk classification on record |
| 3 | **Human Oversight (HITL)** | Article 14 | 20 | High-risk/unacceptable-risk agents are covered by an enabled `require_approval` policy |
| 4 | **Documentation** | Article 11 | 15 | Proportion of agent metadata fields (name, description, framework, model provider, model name) that are filled in, across all agents |
| 5 | **Data Retention** | Article 12 | 20 | Days of audit history retained, scored against the 180-day (6-month) minimum |
| 6 | **FRIA Completion** | Article 27 | 10 | Proportion of high-risk/unacceptable-risk agents with a submitted FRIA |
| 7 | **Incident Response** | Article 73 | 10 | No incidents currently past their Article 73 notification deadline |
## Get Your Readiness Score
```bash curl theme={null}
curl https://api.meshai.dev/api/v1/compliance/readiness \
-H "Authorization: Bearer msh_YOUR_API_KEY"
```
```python SDK theme={null}
from meshai import MeshAI
client = MeshAI(api_key="msh_...")
score = client.get_readiness_score()
```
```json Response theme={null}
{
"success": true,
"data": {
"score": 85,
"max_score": 120,
"components": [
{
"name": "Audit Trail Active",
"article": "Article 12",
"max_score": 20,
"score": 20,
"detail": "234 events recorded",
"status": "pass"
},
{
"name": "Risk Classification",
"article": "Article 6",
"max_score": 25,
"score": 21,
"detail": "10/12 agents classified",
"status": "partial"
},
{
"name": "Human Oversight (HITL)",
"article": "Article 14",
"max_score": 20,
"score": 20,
"detail": "2 approval policies cover 3 high-risk agents",
"status": "pass"
},
{
"name": "Documentation",
"article": "Article 11",
"max_score": 15,
"score": 13,
"detail": "13/15 metadata completeness",
"status": "pass"
},
{
"name": "Data Retention",
"article": "Article 12",
"max_score": 20,
"score": 20,
"detail": "212 days retained",
"status": "pass"
},
{
"name": "FRIA Completion",
"article": "Article 27",
"max_score": 10,
"score": 3,
"detail": "1/3 FRIAs submitted",
"status": "partial"
},
{
"name": "Incident Response",
"article": "Article 73",
"max_score": 10,
"score": 10,
"detail": "No overdue incident notifications",
"status": "pass"
}
],
"assessed_at": "2026-03-17T10:00:00Z"
}
}
```
Each component reports its own `status` - `pass`, `partial`, or `fail` - instead of a single overall rating; there is no separate `total_score`, `rating`, or `recommendations` field. Use `score` / `max_score` at the top level for the overall figure, and inspect each component's `status`/`detail` to see what to fix.
## Improving Your Score
Any governance action (agent registration, kill switch, risk classification, policy changes) writes an audit event. As long as at least one event exists, this component is full.
Assign a [risk level](/governance/risk-classification) to every registered agent. Use the AI-assisted suggestion endpoint if unsure.
Create an enabled `require_approval` [policy](/governance/policies) scoped (via `conditions.risk_levels`) to your high-risk and unacceptable-risk agents.
Ensure every agent has a name, description, framework, model provider, and model name via `PATCH /agents/{id}`.
This component scores against your oldest retained audit event. It rises automatically as your account accumulates history, up to the 6-month (180-day) minimum.
Complete and submit a [FRIA](/governance/fria) for every high-risk or unacceptable-risk agent.
Resolve or notify authorities before the Article 73 deadline on any open [incident](/governance/incidents) - this component fails if any incident is currently overdue.