SQL Guard Copilot — SQL 防护 Copilot

SQL 查询 Copilot Overview

Use this 技能 to turn plAIn-language 请求s into executable SQL with a predictable, low-risk 工作流. Default to read-only execution and 验证 every 查询 agAInst 模式 before 运行ning.

Quick 启动

设置 SQL_DSN first (or pass --dsn each time).

# PowerShell $env:SQL_DSN="mysql://user:password@127.0.0.1:3306/stock_监控" $env:SQL_DSN="postgres://user:password@127.0.0.1:5432/stock_监控" $env:SQL_DSN="sqlite:///d:/data/demo.db"

# Windows CMD 设置 SQL_DSN=mysql://user:password@127.0.0.1:3306/stock_监控 设置 SQL_DSN=postgres://user:password@127.0.0.1:5432/stock_监控 设置 SQL_DSN=sqlite:///d:/data/demo.db

# Bash / Zsh 导出 SQL_DSN="mysql://user:password@127.0.0.1:3306/stock_监控" 导出 SQL_DSN="postgres://user:password@127.0.0.1:5432/stock_监控" 导出 SQL_DSN="sqlite:///d:/data/demo.db"

Core commands:

python scripts/sql_easy.py tables python scripts/sql_easy.py describe dAIly_kline python scripts/sql_easy.py lint --sql "SELECT FROM dAIly_kline" python scripts/sql_easy.py explAIn --sql "SELECT code, close FROM dAIly_kline WHERE trade_date >= '2026-01-01'" python scripts/sql_easy.py 查询 --sql "SELECT code, close FROM dAIly_kline ORDER BY trade_date DESC" --limit 50 python scripts/sql_easy.py 查询 --sql "SELECT code, close FROM dAIly_kline" --summary python scripts/sql_easy.py ask --q "show symbols with old sell 签名als older than 20 days" --summary python scripts/sql_easy.py 性能分析

设置 OPENAI_API_KEY (or pass --API-key) to use ask.

v0.2 Highlights Multi-engine support: MySQL, PostgreSQL, SQLite. SQL lint engine: catches high-risk patterns before execution. ExplAIn mode: quickly inspect 查询 plan (EXPLAIN / EXPLAIN 查询 PLAN). Natural-language mode: ask 生成s SQL from user intent. 查询 summary: auto 性能分析 returned columns (null ratio, distinct count, min/max/avg). Slow 查询 警告: highlights expensive queries using --slow-ms. 审计 记录: write command metadata to JSONL via --审计-记录 or SQL_EASY_审计_记录. 工作流

Clarify the metric and grAIn. Ask for time window, dimensions, and 输出 columns before writing SQL.

Discover 模式 first. 运行 tables, describe , and 性能分析 before any complex SQL.

Draft SQL in read-only mode. Use SELECT or WITH; keep columns explicit and 添加 time 过滤器s.

执行 with 防护rAIls. 运行 via scripts/sql_easy.py 查询, keep --limit unless full 导出 is explicitly needed.

验证 结果s. Cross-检查 row count, null ratio, and edge dates; adjust 查询 and re运行.

防护rAIls Default to read-only SQL. Reject destructive 状态ments (INSERT, 更新, 删除, DROP, ALTER, T运行CATE, etc.). Prefer explicit columns over SELECT for production/报告 queries. 运行 lint before heavy or scheduled queries. 运行 explAIn before 应用roving complex joins/window queries. Always quote identifiers when table/column names are uncertAIn. For business decisions, provide 机器人h SQL and a short interpretation of returned data. 查询 Patterns

Read references/查询_patterns.md when creating:

Top-N and ranking queries Time-window aggregation Dedup with window functions Funnel-style conditional counts Data 质量 检查s (null/duplicate/outlier)

Read references/chanquant_templates.md for Chanquant-specific 查询 templates.