postgresql-best-practices-cloudbase
CloudBase PostgreSQL access-pattern and slow-query quality guidance. Use when designing how tables are read and written, eliminating per-row database calls, adding indexes for filters or joins, reviewing high-traffic data access, or explaining a slow SQL query. Not for first-time PG SDK setup, login UI, or NoSQL collections.
- License 许可
- MIT
- 最近更新
- 2026-11-11
- Stars
- ★ 447
- Forks
- 138
01 前言 Front Matter
- name
- postgresql-best-practices-cloudbase
- license
- MIT
- version
- 2.36.2
- alwaysApply
- false
- description
- CloudBase PostgreSQL access-pattern and slow-query quality guidance. Use when designing how tables are read and written, eliminating per-row database calls, adding indexes for filters or joins, reviewing high-traffic data access, or explaining a slow SQL query. Not for first-time PG SDK setup, login UI, or NoSQL collections.
02 资源文件 4
- 参考文档references/access-patterns.md3.6 KB
- 参考文档references/capacity-and-connections.md3.1 KB
- 参考文档references/indexes-and-explain.md2.6 KB
- 参考文档references/schema-and-rls.md3.2 KB
03 SKILL.md 正文
CloudBase PostgreSQL Best Practices
This skill turns a working CloudBase PG implementation into a reviewable access design. It makes query paths, indexes, row authorization, and launch capacity explicit before code is considered complete.
Sibling skills (local only)
Sibling CloudBase skills ship beside this skill. Use local relative paths such as ../postgresql-development-cloudbase/SKILL.md.
If a referenced sibling skill file is missing from this environment, ask the user to install the full CloudBase plugin (or the missing skill). Do not HTTP-fetch remote skill or protocol markdown into the agent context.
When to use
- A feature needs a new or changed access pattern.
- Application code performs repeated, serial, or high-frequency PG queries.
- A SQL query, RPC, endpoint, or polling path is slow.
- A high-traffic launch needs a database readiness review.
Do NOT use for
- PG environment setup,
app.rdb()syntax, migrations, GRANTs, or base RLS correctness: use../postgresql-development-cloudbase/SKILL.md. - Login provider or UI implementation: use the matching auth skill.
- Confirmed NoSQL collection work: use the document database skill.
- Platform alarm diagnosis or slow-log collection: use
../ops-inspector/SKILL.md.
Workflow
1. Inventory access paths
List every changed endpoint, job, RPC, or UI flow and its reads and writes. Record filters, joins, sort keys, expected cardinality, and request frequency.
Complete when: every changed path has a concrete query inventory; unknown volume or peak concurrency is marked as an assumption for the user.
2. Design schema, indexes, and row authorization together
Derive columns and indexes from the query inventory. Every frequent filter, join, and stable ordering path must have a deliberate index decision. For user-owned rows, use the role as the policy gate and identity as the row predicate:
CREATE POLICY orders_select_own ON public.orders
FOR SELECT TO authenticated
USING ((SELECT auth.uid()) = owner_id);
Apply schema DDL through the versioned applyMigration workflow defined by ../postgresql-development-cloudbase/SKILL.md. This keeps Git, development, and production reproducible; the tradeoff is that experiments require a new migration version instead of ad hoc schema mutation.
Complete when: every query predicate has an index decision, every user-owned path has an ownership predicate, and all DDL is represented in one reviewable migration for the change.
3. Implement bounded database access
Batch related keys, combine repeated reads, and execute independent queries concurrently. Keep each result bounded by a filter plus .limit() / .range(), or by an RPC that implements keyset pagination.
const userIds = [...new Set(rows.map((row) => row.user_id))];
const { data: profiles, error } = await db
.from("profiles")
.select("id, display_name")
.in("id", userIds);
if (error) throw error;
Send request logs, traces, and high-volume analytics to the platform logging service or purpose-built analytics storage. Keep PG for transactional data that participates in business queries and constraints.
Complete when: the changed code has no database call inside an item loop, no repeated read of the same row in one request, and no unbounded hot-path query.
4. Check launch capacity when traffic is material
For campaigns, rankings, polling endpoints, or other bursty paths, inspect the target environment with queryEnv(action="info", envId=...). Report the observed PG allocation, expected peak load, and unresolved capacity risk. CloudBase currently requires an explicit capacity plan; do not represent autoscaling as guaranteed.
Complete when: the current allocation is recorded and either judged against an explicit traffic assumption or raised to the user as an unresolved launch blocker.
Routing
| Current branch | Read |
|---|---|
| Repeated calls, serial queries, polling, pagination, or logging tables | references/access-patterns.md |
| Missing indexes, slow SQL/RPC, or query-plan review | references/indexes-and-explain.md |
| New schema, tenant isolation, or RLS performance | references/schema-and-rls.md |
| Campaign launch or database sizing | references/capacity-and-connections.md |
Load only the references required by the current branch.
Minimum self-check
- Every changed access path is accounted for.
- Every frequent filter, join, and stable sort has an index decision.
- Database round trips are bounded independently of result cardinality.
- RLS separates role gating (
TO) from row authorization (auth.uid()). - Schema changes use the canonical migration workflow.
- Capacity claims are based on observed allocation and stated traffic assumptions.
04 同仓库更多 Skills
yingmi-skill
当用户需要查询基金、策略、公告、财经资讯,做资产配置、组合诊断、风险回测、现金流分析,或生成图表、PDF 时,优先使用本 Skill 获取真实数据与可执行能力。
web-access
所有联网操作必须通过此 skill 处理,包括:搜索、网页抓取、登录后操作、网络交互等。
tencentmap-map-assistant
腾讯位置服务·地图助手 Skill,一句自然语言调用腾讯地图全套能力,无需开发者账号、开箱即用。提供 AI 旅游攻略、地点搜索(含评分/人均/营业时间)、关键词提示、路线规划(驾车/步行/公交/骑行)、地址解析与逆解析、行政区划、IP 定位、距离计算、天气查询,并可将行程或多 POI 渲染成网页地图。涉及找地点、规划路线、旅游行程、查天气、坐标转换等出行场景时使用。
skill-scanner
Scan any agent skill for security risks before you install or use it. Powered by Tencent Zhuque Lab A.I.G (AI-Infra-Guard). 100% local static analysis — no file contents or credentials leave your device. Compatible with CodeBuddy, Cursor, Windsurf, Claude Code, OpenClaw and more. Triggers on: 这个 skill 安全吗, skill 安全扫描, 检查 skill 安全, audit skill, scan skill, check skill safety, analyze skill, inspect skill, verify skill, skill security, skill supply chain. Do NOT trigger for general agent usage, full system health checks, project debugging, or normal development.
skill-creator
Guide for creating effective skills. This skill should be used when users want to create a new skill (or update an existing skill) that extends agent capabilities with specialized knowledge, workflows, or tool integrations.
shader-dev
Comprehensive GLSL shader techniques for creating stunning visual effects — ray marching, SDF modeling, fluid simulation, particle systems, procedural generation, lighting, post-processing, and more.