The Decisions page tracks every choice that matters: things pending a verdict, things already resolved, and things that have been escalated for human approval before action can proceed.
---
## Decision types
| Type | Meaning |
|---|---|
| **pending** | A question or fork that still needs an answer — actively blocking or influencing work |
| **made** | A resolved choice; kept for historical context and to explain why things are the way they are |
---
## Decision statuses
| Status | Border | Meaning |
|---|---|---|
| **open** | blue | Pending, no urgent flag |
| **escalated** | amber | Requires human approval before any action (constitution §4) |
| **resolved** | green | Decision has been made and recorded |
| **superseded** | gray | Replaced by a newer decision |
Decisions are sorted by status (escalated → open → resolved → superseded), then by deadline (earliest first within the same status group).
---
## Resolution History chart
An area + step chart showing cumulative decisions over time. Each dot marks a period where at least one decision was recorded or resolved.
**Period selector** — choose the time window:
| Period | Bucket | Use for |
|---|---|---|
| day | hour | Today's activity |
| week | day | Rolling 7 days |
| month | week | Current calendar month |
| quarter | month | Current quarter |
| YTD | month | Year to date |
| year | month | Rolling 12 months |
| all | month | Full history |
The chart respects the active filter — changing type, status, or search narrows the data plotted.
---
## Filter bar
| Filter | Effect |
|---|---|
| **Type** | Show only `pending` or `made` decisions |
| **Status** | Show one or more of open / escalated / resolved / superseded |
| **Search** | Case-insensitive substring match on the decision title |
All filters are applied client-side on the last 500 decisions fetched from the API.
---
## Decision cards
Each card shows:
- **Type badge** — amber `pending` or indigo `made`
- **Status badge** — colour-coded (see table above)
- **Domain** — which of the six tracked domains this decision belongs to (if linked to a topic)
- **Due date** — shown in red with ⚠ if the deadline has passed
- **Age badge** — `open Xd` for unresolved decisions; `took Xd` for resolved ones. Turns amber when an open decision has been open longer than the current average resolution time.
- **Created date** — when the decision was first recorded
- **Title** — the decision question or statement
- **Description / rationale** — first 200 characters, truncated with `…`
- **Resolved by** — who resolved it and when (resolved decisions only)
- **Escalation note** — amber panel explaining why human approval is required (escalated decisions only)
---
## Decision Health card
The **Decision Health** widget in the right margin shows two KPIs computed from the full (unfiltered) dataset:
| KPI | How it's calculated |
|---|---|
| **avg resolve** | Mean time from creation to resolution, computed over the last 5 resolved decisions |
| **avg open age** | Mean age of all currently open or escalated decisions |
**Color coding for avg open age:**
| Color | Condition |
|---|---|
| black (inherit) | All open decisions younger than the avg resolve time |
| amber | At least one open decision older than the avg resolve time |
| red | Mean open age exceeds the avg resolve time |
---
## Escalation
Any `pending` decision whose title or rationale contains financial or legal keywords is automatically escalated by the API at creation time. Escalated decisions:
- Sort to the top of the list
- Carry an amber escalation note explaining the reason
- Trigger the warning box at the bottom of the page listing all escalated items