Content
83%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 content is an excellent lean reference: dense, executable, pattern-based SQL with no padding, well organized by use case. Its weaknesses are the missing validation steps around the system-configuration template (a batch/destructive change applied blindly) and progressive disclosure that relies on inline tables rather than split reference files.
Suggestions
Add a verification step to the configuration template, e.g. checking current grants ('SELECT grantee, privilege_type FROM information_schema.role_table_grants') before the REVOKE, and confirming settings take effect after pg_reload_conf().
Add a brief diagnostic loop for slow queries: run the pg_stat_statements query, EXPLAIN ANALYZE the top offender, then match against the index decision table.
Move the anti-pattern detection queries or config template into a references/ file (or link the database-reviewer agent's location explicitly) to keep SKILL.md a tighter overview.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is almost entirely dense reference tables and terse, commented SQL snippets — 'Orden del Índice Compuesto', 'Paginación por Cursor -- O(1) vs OFFSET que es O(n)' — with zero padding and no explanation of concepts Claude already knows. Every token earns its place. | 5 / 5 |
Actionability | All guidance is copy-paste-ready executable SQL: index-type decision table, covering/partial index patterns, RLS policy with the auth.uid() wrapping gotcha, SKIP LOCKED queue processing, anti-pattern detection queries, and a full configuration template. Specific examples cover the common cases; the only flexibility ('ajustar según RAM') is explicitly justified. | 5 / 5 |
Workflow Clarity | The 'Cuándo Activar' section gives clear activation triggers, but the configuration template performs batch/system-level changes ('ALTER SYSTEM SET max_connections', 'REVOKE ALL ON SCHEMA public FROM public') with no validation or verification steps (e.g., checking existing grants or testing settings before applying), which caps workflow clarity at 3 per the rubric guidelines. Not 2 because the sections that exist are coherent and well-sequenced. | 3 / 5 |
Progressive Disclosure | Well-organized sections with a clearly signaled one-level-deep pointer for detail ('Para orientación detallada, usa el agente database-reviewer') and a Related section linking sibling skills. Not 5 because the body runs ~145 lines of inline reference material (anti-pattern queries, config template) with no bundle files to split into, and the reference/agent pointers are not formatted as navigable file links. | 4 / 5 |
Total | 17 / 20 Passed |