Content
76%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.
The body is a dense, highly actionable troubleshooting reference with concrete code for every common Supabase problem and a useful diagnosis table. Its weaknesses are the absence of explicit verify-the-fix validation checkpoints in the diagnostic flow and heavy inlining of per-domain content that could be split into more reference files.
Suggestions
Add an explicit verification step to the diagnostic flow (e.g., after applying a fix, re-run the failing query / re-check auth.getUser() to confirm the issue is resolved before considering it closed).
Split the larger inlined sections (Authentication, RLS, Database, Realtime) into per-domain reference files under references/, keeping SKILL.md as a concise diagnosis-and-pointer overview to reduce token load.
Trim redundancy between the inline error-code tables and references/common-errors.md so each error lives in one canonical place.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is lean and mostly code/tables with minimal prose padding — it does not explain what Supabase or basic concepts are. It sits at anchor 4 rather than 5 because the SKILL.md inlines a large volume of per-issue code (~450 lines) and a few inline comments are slightly redundant with the dedicated common-errors.md reference. | 4 / 5 |
Actionability | Every section provides copy-paste-ready JavaScript, SQL, or bash snippets plus a symptom→cause diagnosis table and error-code tables with solutions, covering the common Supabase cases. Placeholder values like 'your_table' are appropriate for a reference, fitting anchor 5's fully executable, common-case coverage. | 5 / 5 |
Workflow Clarity | The 'Quick Diagnosis' table gives a rough symptom→cause flow and each section offers a fix, but there is no explicit sequenced process and, critically, no 'verify the fix resolved the issue' checkpoint. Per the destructive/batch-fix cap, a troubleshooting skill that applies fixes (CREATE POLICY, upserts) without validation steps cannot score above 3. | 3 / 5 |
Progressive Disclosure | Structure is good with clear section headers and a clearly signaled, one-level-deep References link to the real references/common-errors.md file. It is below anchor 5 because substantial per-domain code (auth, RLS, database, realtime) is inlined in SKILL.md rather than split into additional reference files, leaving minor organization gaps. | 4 / 5 |
Total | 16 / 20 Passed |