Content
86%Weight 40%Scale 1-5Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
A highly actionable, well-structured skill: every configuration variant is copy-paste ready, clients are handled specifically, and detail is properly offloaded to clearly labeled reference files. The only weaknesses are minor redundancy between the mermaid flow and the questions section, and verification that stops short of inline error recovery.
Suggestions
Collapse the mermaid Decision Flow and the 'Questions to Ask' section into one representation, and merge the two overlapping Local/Remote table rows to remove contradictory-sounding framing.
Add a short inline verification step for the remote server (e.g., a curl or client-side check against the HTTP endpoint) and a one-line 'if this fails, see troubleshooting.md' pointer after each test to close the feedback loop.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is dominated by directly usable material (JSON templates, commands, file paths) with almost no explanation of concepts Claude already knows. However, the mermaid decision flow duplicates the "Questions to Ask" section, and two Local/Remote table rows ("Required for development... can also connect... for consumption" vs "Works without a dbt platform account but can also connect... for development") overlap confusingly, which is trimmable redundancy. | 4 / 5 |
Actionability | Copy-paste-ready JSON configs for every auth mode (CLI-only, OAuth, token, .env, remote), exact commands ("claude mcp add dbt -s user -- uvx dbt-mcp", "uvx --env-file .env dbt-mcp"), and precise client UI steps with config file locations. The common cases are fully covered with executable guidance. | 5 / 5 |
Workflow Clarity | The sequence (decision questions → prerequisites → templates → client setup → verification) is clear and includes validation checkpoints ("No errors = successful configuration", client verification prompts, "MCP: List Servers"). It falls short of the 5 anchor because there is no inline failure-recovery loop (delegated entirely to references/troubleshooting.md) and no test procedure for the remote server configuration. | 4 / 5 |
Progressive Disclosure | Three real, one-level-deep reference files (finding-credentials.md, troubleshooting.md, environment-variables.md) are clearly signaled with purpose statements ("for detailed guidance on obtaining tokens and IDs", "for common issues and fixes"), and detail content is appropriately split out while the core configs stay inline. No nested references; navigation is easy. | 5 / 5 |
Total | 18 / 20 Passed |