ONBOARDDECK

From Markdown to Dashboard in Five Steps

No configuration. No pipeline to maintain. Claude Code already generates the report — OnboardDeck consumes it and turns it into queryable, historically-tracked data.

Every Ingest. Full History.

The main OnboardDeck view shows a reverse-chronological table of all team snapshots. Each row captures the team name, ingest date, number of metrics extracted, checklist items parsed, and top command identified. Click any row to drill into the full parsed detail view. Snapshots are immutable — every ingest adds a new record, never overwriting the past.

Aggregate stats across the header — total snapshots ingested, distinct teams tracked, total metrics extracted, and checklist items accumulated — give you an instant read on how actively your team is engaging with the tool.

Paste. Parse. Done.

The ingest view accepts raw markdown output from your /team-onboarding skill. Paste directly into the text area or upload a .md file (max 2 MB). Add a descriptive label — such as the team name and month — and hit submit.

The tolerant parser handles any valid output format: percentage bars, count values, per-month frequencies, slash commands, and plain text entries. Unrecognised lines are silently ignored — the parser never crashes, no matter how the source markdown is structured.

Deep Dive Into Any Snapshot

Click any snapshot row to open the full detail view. Work-type percentages render as horizontal CSS bar charts — no JavaScript charting libraries required. Command frequencies show use counts and normalized bars side by side. MCP server lists appear as a compact grid.

The right panel shows the raw markdown the snapshot was ingested from, so you always have a verifiable source of truth. The detail view also shows the full team tips section and links directly to the checklist panel for that snapshot.

Track Onboarding Progress Per Snapshot

Every - [ ] item from your team onboarding doc becomes a trackable checklist entry in OnboardDeck. Items are categorised automatically — Codebases to explore, MCP Servers to set up, Skills to learn — based on the markdown section they appear under.

Check off items directly from the snapshot view. Progress persists per snapshot — so if you re-ingest monthly, each month's snapshot maintains its own independent checklist state. Compare completion rates across snapshots to see whether the team is keeping pace with new recommendations.

Watch Usage Evolve Over Time

The trends view aggregates metric values across all snapshots and renders inline SVG line charts for any metric with more than one data point — no JavaScript charting libraries required. Select any metric from the dropdown to plot its value across every ingest date.

A summary table below the chart lists every snapshot's metric values side by side, making it easy to spot step-changes. As your team re-ingests each sprint cycle, the trends view becomes a de facto adoption velocity report.

Everything Included

OnboardDeck ships as a complete first-class blueprint — ingestion, storage, parsing, dashboard, and trends all out of the box. No extra setup required beyond your MelTuc account.

What the Parser Extracts

The OnboardDeck parser is written to handle the exact markdown format produced by Claude Code's /team-onboarding skill. It is deliberately tolerant — if a section is missing, it returns an empty list for that category rather than erroring. You can ingest docs from different team skill versions without any issues.