Desenvolvimento

Auto-activate for Litestar plugins=, InitPluginProtocol, CLIPluginProtocol, SerializationPluginProtocol, OpenAPISchemaPluginProtocol, DomainPlugin, app plugin lists, or first-party plugin composition. Use when wiring Litestar plugins or authoring plugin protocols. Not for installing unrelated Python packages or deployment plugin marketplaces.

Use when user invokes /harness-compile or wants to compile a harness.yaml into native config files for Claude Code, Cursor, and GitHub Copilot. Generates CLAUDE.md, AGENT.md, .mcp.json, .cursor/rules/, .vscode/mcp.json, and related files from a single harness.yaml source. Supports --target, --dry-run, and --clean flags.

Auto-activate for uv build, hatch build, pyapp, Hatchling force-include, bundled frontend assets, PyApp onefile binaries, GitHub Actions build matrices, PYAPP_* env vars, cargo-zigbuild, or python-build-standalone. Use when packaging Litestar apps as wheels, onefile binaries, release artifacts, or asset-bundled distributions. Not for runtime deployment - use litestar-deployment.

Curated edge case test data for boundary testing, verification scripts, and test generation. Provides pre-curated reference data organized by data type with context-specific loading guidance.

Auto-activate for guards=, Guard, ASGIConnection, BaseRouteHandler, connection.user, PermissionDeniedException, JWT auth, OAuth2, role checks, tenant checks, or WebSocket auth. Use when implementing Litestar authentication, authorization, Guards, or permission boundaries. Not for business validation or frontend route protection.

Use when wrapping up a development task and getting a PR ready — runs local tests, creates a PR (if one doesn't exist) with a structured description template, conducts a code review via the review skill, and checks CI status with quick fixes. Trigger when the user says they're done with a feature, want to open a PR, wrap up, finalize their work, or push their branch for review. Also invoke proacti

Auto-activate for litestar.toml, [tool.litestar], pyproject.toml with a litestar dependency, or app creation with Litestar(). Use when orienting a Litestar project, choosing the right Litestar subskill, or scaffolding app-level structure. Not for focused routing, DTO, auth, DI, data, settings, exceptions, middleware, plugin, realtime, AI, deployment, or testing work - use the narrower skill.

Auto-activate for get_env, dataclass settings, pydantic_settings, BaseSettings, lru_cache settings factories, app.state settings, LITESTAR_APP, env var parsing, or typed app config. Use when defining Litestar settings, config factories, env loading, or settings injection. Not for frontend environment variables or deployment secrets management alone.

Challenges the user's reasoning, assumptions, or proposals — a thinking sparring partner that pokes holes without the full critic workflow. Use when the user invokes /roast to stress-test their logic, or /roast <argument> to challenge a specific claim or decision. Not a review of artefacts — a challenge to thinking.

Auto-activate for Google ADK, LlmAgent, Runner, SQLSpecSessionService, Vertex AI, agent chat endpoints, streaming AI responses, tool-calling endpoints, or Litestar-hosted model workflows. Use when serving AI agents or model-backed workflows from Litestar. Not for offline ML training or non-Litestar agent servers.

Auto-activate for litestar.toml, [tool.litestar], pyproject.toml with a litestar dependency, or app creation with Litestar(). Use when orienting a Litestar project, choosing the right Litestar subskill, or scaffolding app-level structure. Not for focused routing, DTO, auth, DI, data, settings, exceptions, middleware, plugin, realtime, AI, deployment, or testing work - use the narrower skill.

<!-- @include templates/prompts/iron-law.md --> 对指定服务或应用进行性能分析并提出优化方案。 数据库层: ? ?查询、N+1 问题、缺失索引、? ?表扫描 缓存层: 缓存命中率、缓存策略（过期/淘汰）、缓存穿透/击穿/雪崩 接口层: 响应时间、并发处理、连接池? ?置 代码层: 算法复杂度、? 存泄漏、不? 要的序列化/反序列化 加载性能: 首屏时间、资源体积、代码分割、懒加载…