Database Performance Audit

Download SKILL.mdGet Murage How skills work

Open source. macOS, Windows and Linux.

At a glance

What it is

Database performance assessment evaluating query efficiency, indexing strategy, connection management, configuration, and scaling readiness to produce an actionable performance scorecard.

A skill is a written procedure a bot loads when a job calls for it. This one is a single file, SKILL.md, and the whole file is on this page.

File: SKILL.md
Length: 350 lines2,243 words, about 11 min read
Category: Backend systemsdatabase
Level: Beginner
License: Apache-2.0

More backend systems skills.

The full skill.

The whole file, exactly as it downloads. Read it before you rely on it.

SKILL.md

350 lines · 2,243 words

Download

Database Performance Audit

You are a senior database architect specializing in performance auditing. Your role is to systematically evaluate database systems across query performance, indexing, connection management, configuration, and operational readiness to produce a structured performance scorecard. You find the bottlenecks that slow applications and the risks that cause outages.

When to Use

Use this skill when:

User asks about database performance audit techniques or best practices

Put this skill to work.

Questions.

What is the Database Performance Audit skill?

Database performance assessment evaluating query efficiency, indexing strategy, connection management, configuration, and scaling readiness to produce an actionable performance scorecard.

How much does it cost?

Nothing. It ships with Murage, which is open source. Your bots run on the AI plan you already have.

Can I change it?

Yes. Open it in Settings → Skills to read, edit or duplicate it. Skill Guard scans it again after every edit.

Give your first job to Murage.

Download the app, connect the AI you already pay for, and tell your Chief of Staff what needs doing. Plan on about ten minutes from install to a working team.

Download Murage

No account needed · Runs on the AI plan you already pay for

Score	Criteria
1	Numerous queries >5s. No query monitoring. Full table scans on large tables. N+1 queries throughout. Application timeouts frequent.
2	Some slow queries identified. Basic monitoring exists. Full table scans on some queries. Average response >500ms.
3	Most queries <200ms. Slow query logging enabled. Known slow queries have optimization plans. Occasional spikes.
4	P95 query time <100ms. Continuous query monitoring. Slow queries addressed proactively. Query plans reviewed regularly.
5	P99 query time <50ms. Automated slow query detection. Query performance regression testing. All queries are optimal for their use case.

Score	Criteria
1	Only primary key indexes. No composite indexes. Missing indexes on foreign keys. No awareness of index usage.
2	Some indexes exist but are ad hoc. Duplicate indexes present. Missing indexes on key query patterns. No index maintenance.
3	Indexes cover primary query patterns. Foreign keys indexed. Some composite indexes. Periodic index review.
4	Comprehensive indexing strategy. Covering indexes for hot queries. Unused indexes identified and removed. Partial indexes used where appropriate.
5	Optimal indexing. Every query uses appropriate indexes. Index bloat managed. Automated index recommendations monitored. Expression indexes where beneficial.

Score	Criteria
1	No connection pooling. Application opens new connections per query. Connection limit regularly hit. Connection storms crash the database.
2	Basic pooling in application. Pool sizes not tuned. Occasional connection exhaustion. No connection monitoring.
3	Connection pooler in place. Pool sizes reasonably configured. Connection utilization monitored. Rare exhaustion events.
4	Optimized pool configuration. External connection pooler (PgBouncer/ProxySQL). Connection limits per tenant/service. Health check connections.
5	Sophisticated connection management. Dynamic pool sizing. Connection routing (read/write splitting). Zero connection-related issues. Prepared statement optimization.

Score	Criteria
1	No normalization strategy. Massive tables with hundreds of columns. No constraints. Data types oversized. No naming conventions.
2	Basic normalization. Some constraints. Mixed naming conventions. Some oversized data types.
3	Appropriate normalization. Constraints enforced. Consistent naming. Proper data types. Foreign keys defined.
4	Well-designed schema. Strategic denormalization for performance. Partitioning for large tables. Check constraints. Domain types.
5	Exemplary schema design. Optimal normalization balance. Table partitioning strategy. Schema versioning. Data lifecycle managed. Archive strategy.

Score	Criteria
1	Default configuration. No tuning attempted. Memory allocation not optimized. Default connection limits.
2	Some basic tuning. Memory settings adjusted. Still many defaults. No workload-specific tuning.
3	Key parameters tuned for workload. Memory allocation optimized. Connection limits set. WAL configuration reasonable.
4	Comprehensive tuning. Workload-specific configuration. Vacuum/autovacuum tuned. Checkpoint configuration optimized. Regular tuning reviews.
5	Expertly tuned. Every significant parameter optimized. Tuning validated with benchmarks. Configuration version controlled. Automatic tuning adjustments.

Score	Criteria
1	No backups. No recovery plan. No point-in-time recovery. A database failure means data loss.
2	Daily backups. Untested restores. No point-in-time recovery. RTO and RPO undefined.
3	Regular backups. Restore tested quarterly. Point-in-time recovery capability. RTO/RPO defined and achievable.
4	Continuous backup with streaming replication. Automated restore testing. PITR to any second. Documented recovery procedures.
5	Multi-region backup strategy. Automated failover. Sub-minute RPO. Recovery time <5 minutes. Regular DR drills. Backup integrity verification.

Score	Criteria
1	No database monitoring. Problems discovered when application errors spike. No visibility into database health.
2	Basic uptime and disk space monitoring. CPU/memory alerts. No query-level monitoring.
3	Query performance monitoring. Connection pool metrics. Replication lag tracking. Meaningful alerts with thresholds.
4	Comprehensive monitoring covering queries, locks, vacuum, replication, disk, connections. Anomaly detection. Historical analysis.
5	Full database observability. Automated query analysis. Predictive capacity alerting. Performance regression detection. Cost optimization insights.

Score Range	Performance Level	Interpretation
4.5 - 5.0	Excellent	Database is well-optimized. Focus on maintaining performance as data grows.
3.5 - 4.4	Good	Solid performance. Address specific gaps for optimal throughput.
2.5 - 3.4	Fair	Performance issues exist. Targeted optimization will yield significant gains.
1.5 - 2.4	Poor	Database is a bottleneck. Significant tuning and architectural changes needed.
1.0 - 1.4	Critical	Database at risk of failure. Emergency intervention required.

Database Performance Audit

More backend systems skills.

The full skill.

Database Performance Audit

When to Use

Put this skill to work.

Questions.

Give your first job to Murage.

Search murage.ai

Database Performance Audit

More backend systems skills.

The full skill.

Database Performance Audit

When to Use

Put this skill to work.

Questions.

Give your first job to Murage.

Create a bot and use this skill.

Ask Ember to create the bot

Open the bot’s skills

Find Database Performance Audit and switch it on

Ask the bot for the job

About this skill.

Questions to Ask First

Database Context

Workload Context

Infrastructure Context

Operational Context

Assessment Framework

Dimension 1: Query Performance (Weight: 25%)

What to Analyze

Dimension 2: Indexing Strategy (Weight: 20%)

What to Evaluate

Dimension 3: Connection Management (Weight: 15%)

What to Evaluate

Dimension 4: Schema Design (Weight: 10%)

What to Evaluate

Dimension 5: Configuration and Tuning (Weight: 10%)

Key Parameters to Review (PostgreSQL Example)

Dimension 6: Backup and Recovery (Weight: 10%)

Dimension 7: Monitoring and Alerting (Weight: 10%)

Key Metrics to Monitor

Scoring Template

Results Interpretation

Recommendations by Priority

Emergency (Today)

Quick Wins (This Week)

Medium-Term (This Month)

Strategic (This Quarter)

Report Template

Process

Output Format

Edge Cases

Example