The Wazzi Metrics Connector

Path: Connectors → Wazzi Metrics → Configuration

What this connector does

Wazzi Metrics is a first-party MCP connector that exposes the metrics platform — categories, collections, views, targets, writers, and weekly values — as MCP tools. It's the successor to the old Scorecard connector, reshaped around the platform's current primitives: categories group related metrics (Growth, Retention, Ops), collections roll metrics up across dimensions, views persist per-audience layouts, and each metric has one or more writers that produce its values on a schedule. This is the write surface behind the in-app Metrics editor.

Because Wazzi Metrics is first-party, there's nothing to configure and no external credentials to enter. Every request is org-scoped server-side, and the connector talks directly to your organization's metrics tables — no third-party API in the loop.

The connector exposes 25 tools grouped by domain:

• Discovery / read (6 tools) — get_status, get_metrics, get_metric_history, list_categories, list_collections, list_views, list_targets, list_metric_owners.
• Metrics — define + write (7 tools) — create_metric, create_metric_set (bulk create), update_metric, delete_metric, set_metric_writer, set_target, set_values.
• Categories (3 tools) — create_category, update_category, delete_category.
• Collections (4 tools) — create_collection, update_collection, delete_collection, assign_metrics_to_collection.
• Views (3 tools) — create_view, update_view, delete_view.
• Attach / detach (2 tools) — attach_metrics, detach_metric.

Before you start

For this connector to have anything to read or write, your organization must have the Metrics platform enabled and at least one category configured. New orgs typically start by creating one or two categories (e.g. Growth, Ops), attaching metrics to each, and setting targets. Once the shape is in place, everything below has data to operate on.

Beyond that, there's no per-user setup — just enable the connector on the user's group in Wazzi.

What this connector unlocks for your team

Once configured, members with the right MCP permissions can ask Claude / ChatGPT / Cursor things like:

• "List every metric in the Growth category and show me their current value vs target."
• "Pull the last 12 weeks of values for the guest_nps metric and summarize the trend."
• "Create a new metric called qa_score in the Ops category — percentage, higher-is-better, weekly grain — and set an initial target of >=95%."
• "Add the sarah_utilization metric to the Ops collection so it rolls up with the rest of the team's utilization."
• "Create a view called 'Weekly Exec Review' with the Portfolio-level view of every metric in the Growth and Retention categories."
• "Set Sarah Chen as the writer for the guest_nps metric — she'll be prompted to submit the weekly value."
• "Set target on ADR to $275 >= for the Portfolio dimension with the note 'Q1 2026 board goal.'"
• "Manually set this week's booked_nights to 412 on Phoenix — the auto-collector missed a source and we've verified the true value."
• "Assign the retention_30d metric to the Retention collection with a note explaining why."
• "Rename the Growth category to Growth & Acquisition."

The exact set of available actions depends on which MCP toggles are flipped on for the user's group in Wazzi — see Managing Permissions. Write tools (create / update / delete / set) can be restricted independently from reads, so an analyst group can query while an ops-lead group can also mutate.

Steps — enabling Wazzi Metrics in Wazzi

1. From the Connectors catalog, click Configure on the Wazzi Metrics tile.

You'll land on Wazzi Metrics' Configuration tab:

2. Read the notice — there are no credentials to enter.

The configuration tab shows an information notice reminding you that the Metrics platform must be enabled and at least one category configured for the tools to have data to operate on.

3. Click Test Connection.

Wazzi runs a quick count query against your metrics tables. On success you'll see a summary of what's configured (metric / category / collection counts) or a nudge to create your first category if the platform is empty.

4. Click Save Configuration.

Wazzi Metrics moves to the Active section of the catalog.

Troubleshooting

• list_categories returns empty. Your org hasn't created any categories yet. Create one with create_category (or in the Metrics UI) before attaching metrics.
• get_metrics returns rows with null current_value. The metric has no writer assigned or the writer hasn't produced a value for the current period. Assign a writer with set_metric_writer, or write a manual value with set_values.
• set_values returns "Past-period writes are locked". Only the current period is writable by default. Pass force: true to override a closed period — the same "Unlock" action the UI exposes.
• set_target fails with "target_value_high is required". Range operators (in_range, not_in_range) need both bounds; target_value_high must be >= target_value.
• delete_category refuses because a metric is still attached. Detach every metric with detach_metric (or move them to another category via update_metric) before deleting.
• create_metric with an id returns 409 conflict. A metric with that id already exists for the org. Either pick a different id or call update_metric to patch the existing one. Ids must be lowercase snake_case.
• Values I set manually get overwritten by the next collection run. The writer's next run overwrote your manual value. Either mark the writer inactive with set_metric_writer, or accept that manual writes are transient for auto-collected metrics.

Best practices

• Group with categories, roll up with collections. Categories are visual/UX grouping (Growth / Retention / Ops). Collections are analytic grouping — how metrics aggregate across dimensions. Use both.
• Use create_metric_set for bulk installs. Migrating from a spreadsheet? One call with the full metric list beats N sequential creates and preserves ordering.
• Persist your executive reviews as views. A saved view (create_view) locks in which metrics + dimensions render on that dashboard, so the same layout re-serves next week without hand-picking.
• Assign a writer for every metric. Metrics without writers stay null forever. set_metric_writer either points at a person (BambooHR id) or at an automated collector.
• Restrict destructive tools. delete_metric, delete_category, delete_collection, and delete_view are broad-reach operations. Only grant those permissions to analytics leads — see Managing Permissions.

Frequently asked questions

Do I need to configure anything?
No. This is a first-party connector with no external credentials. Just enable it on the user's group in Wazzi.

Can users from another organization see or edit my metrics?
No. Every tool call filters by organization_id — metrics, categories, collections, views, targets, and weekly values are all org-scoped. Cross-org access is impossible.

What's the difference between a category and a collection?
A category is a UX grouping — where a metric appears in the Metrics editor's left rail (Growth, Retention, Ops). A collection is an analytic grouping — how metrics roll up across dimensions on a dashboard. A metric belongs to exactly one category but can belong to many collections.

What's a view?
A saved layout — pick a set of metrics + dimensions and save them together so the same shape re-renders when a member opens the dashboard. Views can be personal or shared, and are how audience-specific dashboards (Exec review, Ops standup, etc.) are persisted.

Isn't this the same thing as the old Scorecard connector?
Roughly the same use cases, but reshaped. The old Scorecard had dimensions and sections as first-class layout primitives; the current Metrics platform replaces them with views for layout and categories + collections for grouping. Different tool names, richer surface, same "what number are we hitting this week" outcome.

Why can I only write to the current period by default?
Locking past periods prevents accidental history rewrites — the metrics platform's whole value is showing trend over time, and unrestricted past writes erode trust in the numbers. Set force: true on set_values when you truly need to fix a past period.

What's next

• Set up your first categories and metrics in the Metrics editor.
• Restrict who in your org can invoke Wazzi Metrics actions: Managing Permissions.
• Wire Wazzi Metrics into your AI tool: Connecting Claude or Connecting ChatGPT.