# Dashboard Data Contract

Use `schemas/dashboard-data.schema.json` as the canonical validation contract.

The schema accepts two dashboard data shapes:

1. **Compact data mode** for agents that emit `app_name`, `findings`,
   `remediation_waves`, and `learning_priorities`.
2. **HTML-template mode** for the bundled templates that use `project`,
   `backlog`, and `learning_plan`.

Prefer compact data mode for new dashboard data. Use HTML-template mode when
replacing the JSON inside an existing template.

```yaml
app_name: ""
score: 0
confidence: "low|medium|high"
detected_stack: []
summary: ""
counts:
  P0: 0
  P1: 0
  P2: 0
  P3: 0
coverage:
  - surface: "Auth and permissions"
    status: "Pass|Finding|Unknown|N/A"
    evidence: "short evidence"
findings:
  - id: "P1-001"
    severity: "P1"
    title: ""
    plain_risk: ""
    evidence: ""
    status: "open|proposed|approved|fixed|accepted-risk|deferred|not-applicable|suppressed"
remediation_waves:
  - wave: "Wave 1"
    goal: ""
    items: []
next_approval: ""
learning_priorities:
  - concept: ""
    plain_english: ""
    plain_secondary_language: ""
    secondary_language: ""
    why: ""
    triggered_by: []
    do_this_next: ""
    do_this_next_secondary_language: ""
    success_signal: ""
    source_title: ""
    source_url: ""
    source_type: ""
    authority_level: "high|medium|supplemental"
    why_this_source_is_trusted: ""
    checked_at: "YYYY-MM-DD"
    youtube_title: ""
    youtube_url: ""
    video_trust_note: ""
    video_authority_level: "high|medium|supplemental"
```

Language fields should be selected at runtime. Use
`plain_secondary_language` and `do_this_next_secondary_language` only after the
user explicitly requests a second language or confirms an inferred candidate.
The older `plain_other_language` and `do_this_next_other_language` fields are
accepted for compatibility, but new outputs should prefer the runtime-neutral
field names above.