AgentSkillsCN

dev-workflow

SignalDB开发工作流——包括构建、测试、Lint检查、代码格式化、服务运行、Docker容器支持、Grafana插件集成、健康检查，以及语义化提交规范。适用于SignalDB的构建、测试、运行或部署场景。

SKILL.md
--- frontmatter
name: dev-workflow
description: SignalDB development workflow - build, test, lint, format, run services, Docker, Grafana plugin, health checks, and semantic commits. Use when building, testing, running, or deploying SignalDB.

SignalDB Development Workflow

Build

bash
cargo build                # Debug build
cargo build --release      # Release build
cargo build -p <package>   # Build specific crate

Test

bash
cargo test                          # All tests
cargo test -p <package>             # Specific crate (common, writer, acceptor, etc.)
cargo test <test_name>              # Specific test by name
cargo test -- --nocapture           # With stdout visible
RUST_LOG=debug cargo test <name> -- --nocapture   # With logging
cargo test -p tests-integration     # Integration tests

Pre-Commit Checks (MANDATORY before committing)

These run automatically via cargo-husky hooks, but run manually to catch issues early:

bash
cargo fmt                                               # Format code
cargo clippy --workspace --all-targets --all-features   # Lint
cargo machete --with-metadata                           # Unused dependencies
cargo deny check                                        # License/security audit

Running Services

Monolithic Mode

bash
cargo run --bin signaldb                              # All services in one process
./scripts/run-dev.sh                                  # With local file storage
./scripts/run-dev.sh --sqlite                         # SQLite (default, no deps)
./scripts/run-dev.sh --with-deps --postgres           # With PostgreSQL via docker compose

Microservices Mode

bash
cargo run --bin acceptor   # OTLP ingestion (:4317/:4318)
cargo run --bin router     # HTTP router (:3000) + Flight (:50053)
cargo run --bin writer     # Data persistence (Flight :50061)
cargo run --bin querier    # Query execution (Flight :50054)
bash
./scripts/run-dev.sh services   # All microservices (logs to .data/logs/)

Storage Locations

  • •WAL: .data/wal/
  • •Parquet data: .data/storage/
  • •SQLite databases: .data/*.db

Docker

bash
docker compose up          # Start PostgreSQL, Grafana, MinIO, SignalDB
docker compose up --build  # Build images first
docker compose build       # Build only

MinIO: Console localhost:9001, API localhost:9000 (minioadmin/minioadmin)

Grafana Plugin

bash
npm install                     # From workspace root
npm run grafana:dev             # Watch + rebuild frontend
npm run grafana:build           # Production build
cd src/grafana-plugin && npm run build:backend   # Rust backend

Health Checks

bash
curl http://localhost:4318/health   # Acceptor
curl http://localhost:3000/health   # Router

Semantic Commits

Use semantic commit messages:

  • •feat: new feature
  • •fix: bug fix
  • •refactor: code restructuring
  • •docs: documentation
  • •test: test changes
  • •chore: build/tooling changes

Configuration

Precedence: defaults -> signaldb.toml -> env vars (SIGNALDB_*)

bash
cp signaldb.dist.toml signaldb.toml   # Create local config

Key env vars for S3/MinIO:

bash
AWS_ENDPOINT_URL=http://localhost:9000
AWS_ACCESS_KEY_ID=minioadmin
AWS_SECRET_ACCESS_KEY=minioadmin
AWS_REGION=us-east-1