fde-dashboardlisted

# @fde-dashboard ## Token efficiency Load per engagement: `reality.md`, `brief.md`, `success.md`, `risks.md`, `decisions.md`, `delivery.md`, and `stakeholders.md`. Do not load `terrain.md` (too large) or `trust-profile.md` (contains regulated data -- sensitive). `stakeholders.md` is needed specifically for trust signal detection: it is the file where signals like "stakeholder going quiet" or "routing around the FDE" are recorded, and these are the early warning signs that matter most across a portfolio. ## Purpose An FDE works across multiple engagements simultaneously. This skill reads all `.fde/` directories it can find and generates a single clean HTML file, a reflection board showing the full picture across every project. ## What it does Scan engagement folders: 1. **Primary:** `~/fde-engagements/*/.fde/` (default — one folder per engagement) 2. Workspace `./.fde/` if present (optional, engagement-approved) 3. Paths the FDE names explicitly **Multiple engagements:** `npx fdeos init <name>` per engagement under `~/fde-engagements/`. Workspaces are separate. Do not merge two engagements into one `.fde/`. Read each separately. Generate `fde-dashboard.html` in the current directory. Open in a browser. That's it. ## What each engagement card shows - Engagement name and client/context label - Current phase and week - The real problem (from `reality.md`) vs the original brief (from `brief.md`) - Definition of success (from `success.md`) - Top active risk (from `risks.m