Tableau Pulse Blueprint
A Codex skill for designing, reviewing, and optimizing Tableau Pulse metric definitions from business questions, datasource metadata, and existing configurations.
What it provides
- decision-centered metric blueprints;
- ranked metric-program recommendations;
- evidence-aware audits and optimization plans;
- current-schema handling for Pulse enums and MCP capabilities;
- explicit approval boundaries for creates, updates, deletes, subscriptions, and notifications;
- deterministic audit scoring with partial-coverage protection.
Example prompts
Design a Tableau Pulse metric for monthly recurring revenue.Recommend Pulse metrics for this pipeline datasource.Audit my accessible Pulse metric definitions.Optimize our churn metric for weekly operating reviews.Compare this Pulse audit with my previous baseline.Apply the previously agreed formatting changes to these Pulse definition IDs.
Package layout
| Path | Purpose |
|---|---|
SKILL.md | Core routing, workflow, and mutation boundaries |
skill-expertise/tableau-pulse-capabilities.md | MCP/API capability and enum handling |
skill-expertise/metric-design.md | Measure, grain, dimension, insight, and filtered-metric design |
skill-expertise/audit-methodology.md | Contextual audit rules and scoring |
skill-expertise/audit-scoring-contract.md | Normalized helper schema |
skill-expertise/output-contract.md | Blueprint, program, audit, and optimization formats |
skill-expertise/state-and-comparison.md | Consent-based baselines and monitoring |
skill-expertise/score_audit.py | Deterministic audit scoring engine |
skill-expertise/test_score_audit.py | Success, failure, coverage, and portability tests |
Local validation
python -m unittest discover -s skill-expertise -v
python -m py_compile skill-expertise/score_audit.py
Run scoring from any directory with absolute paths:
python /absolute/path/to/pulse-blueprint/skill-expertise/score_audit.py /absolute/path/to/audit.json --pretty
The skill does not assume that a particular Tableau MCP is read-only or writable. It discovers current capabilities and honors explicit authorization already given for bounded changes and asks only when scope or impact remains unclear.