ReddRadar
Agent skill
evals-context
Provides context about the CoStrict evals system structure in this monorepo. Use when tasks mention "evals", "evaluation", "eval runs", "eval exercises", or working with the evals infrastructure. Helps distinguish between the evals execution system (packages/evals, apps/web-evals) and the public website evals display page (apps/web-roo-code/src/app/evals).
Install this agent skill to your Project
npx add-skill https://github.com/zgsm-ai/costrict/tree/main/.roo/skills/evals-context
SKILL.md
Evals Codebase Context
When to Use This Skill
Use this skill when the task involves:
- Modifying or debugging the evals execution infrastructure
- Adding new eval exercises or languages
- Working with the evals web interface (apps/web-evals)
- Modifying the public evals display page on roocode.com
- Understanding where evals code lives in this monorepo
When NOT to Use This Skill
Do NOT use this skill when:
- Working on unrelated parts of the codebase (extension, webview-ui, etc.)
- The task is purely about the VS Code extension's core functionality
- Working on the main website pages that don't involve evals
Key Disambiguation: Two "Evals" Locations
This monorepo has two distinct evals-related locations that can cause confusion:
| Component | Path | Purpose |
|---|---|---|
| Evals Execution System | packages/evals/ |
Core eval infrastructure: CLI, DB schema, Docker configs |
| Evals Management UI | apps/web-evals/ |
Next.js app for creating/monitoring eval runs (localhost:3446) |
| Website Evals Page | apps/web-roo-code/src/app/evals/ |
Public roocode.com page displaying eval results |
| External Exercises Repo | Roo-Code-Evals | Actual coding exercises (NOT in this monorepo) |
Directory Structure Reference
packages/evals/ - Core Evals Package
packages/evals/
├── ARCHITECTURE.md # Detailed architecture documentation
├── ADDING-EVALS.md # Guide for adding new exercises/languages
├── README.md # Setup and running instructions
├── docker-compose.yml # Container orchestration
├── Dockerfile.runner # Runner container definition
├── Dockerfile.web # Web app container
├── drizzle.config.ts # Database ORM config
├── src/
│ ├── index.ts # Package exports
│ ├── cli/ # CLI commands for running evals
│ │ ├── runEvals.ts # Orchestrates complete eval runs
│ │ ├── runTask.ts # Executes individual tasks in containers
│ │ ├── runUnitTest.ts # Validates task completion via tests
│ │ └── redis.ts # Redis pub/sub integration
│ ├── db/
│ │ ├── schema.ts # Database schema (runs, tasks)
│ │ ├── queries/ # Database query functions
│ │ └── migrations/ # SQL migrations
│ └── exercises/
│ └── index.ts # Exercise loading utilities
└── scripts/
└── setup.sh # Local macOS setup script
apps/web-evals/ - Evals Management Web App
apps/web-evals/
├── src/
│ ├── app/
│ │ ├── page.tsx # Home page (runs list)
│ │ ├── runs/
│ │ │ ├── new/ # Create new eval run
│ │ │ └── [id]/ # View specific run status
│ │ └── api/runs/ # SSE streaming endpoint
│ ├── actions/ # Server actions
│ │ ├── runs.ts # Run CRUD operations
│ │ ├── tasks.ts # Task queries
│ │ ├── exercises.ts # Exercise listing
│ │ └── heartbeat.ts # Controller health checks
│ ├── hooks/ # React hooks (SSE, models, etc.)
│ └── lib/ # Utilities and schemas
apps/web-roo-code/src/app/evals/ - Public Website Evals Page
apps/web-roo-code/src/app/evals/
├── page.tsx # Fetches and displays public eval results
├── evals.tsx # Main evals display component
├── plot.tsx # Visualization component
└── types.ts # EvalRun type (extends packages/evals types)
This page displays eval results on the public roocode.com website. It imports types from @roo-code/evals but does NOT run evals.
Architecture Overview
The evals system is a distributed evaluation platform that runs AI coding tasks in isolated VS Code environments:
┌─────────────────────────────────────────────────────────────┐
│ Web App (apps/web-evals) ──────────────────────────────── │
│ │ │
│ ▼ │
│ PostgreSQL ◄────► Controller Container │
│ │ │ │
│ ▼ ▼ │
│ Redis ◄───► Runner Containers (1-25 parallel) │
└─────────────────────────────────────────────────────────────┘
Key components:
- Controller: Orchestrates eval runs, spawns runners, manages task queue (p-queue)
- Runner: Isolated Docker container with VS Code + CoStrict extension + language runtimes
- Redis: Pub/sub for real-time events (NOT task queuing)
- PostgreSQL: Stores runs, tasks, metrics
Common Tasks Quick Reference
Adding a New Eval Exercise
- Add exercise to Roo-Code-Evals repo (external)
- See
packages/evals/ADDING-EVALS.mdfor structure
Modifying Eval CLI Behavior
Edit files in packages/evals/src/cli/:
runEvals.ts- Run orchestrationrunTask.ts- Task executionrunUnitTest.ts- Test validation
Modifying the Evals Web Interface
Edit files in apps/web-evals/src/:
app/runs/new/new-run.tsx- New run formactions/runs.ts- Run server actions
Modifying the Public Evals Display Page
Edit files in apps/web-roo-code/src/app/evals/:
evals.tsx- Display componentplot.tsx- Charts
Database Schema Changes
- Edit
packages/evals/src/db/schema.ts - Generate migration:
cd packages/evals && pnpm drizzle-kit generate - Apply migration:
pnpm drizzle-kit migrate
Running Evals Locally
# From repo root
pnpm evals
# Opens web UI at http://localhost:3446
Ports (defaults):
- PostgreSQL: 5433
- Redis: 6380
- Web: 3446
Testing
# packages/evals tests
cd packages/evals && npx vitest run
# apps/web-evals tests
cd apps/web-evals && npx vitest run
Key Types/Exports from @roo-code/evals
The package exports are defined in packages/evals/src/index.ts:
- Database queries:
getRuns,getTasks,getTaskMetrics, etc. - Schema types:
Run,Task,TaskMetrics - Used by both
apps/web-evalsandapps/web-roo-code
Recommended Agent Skills
Expand your agent's capabilities with these related and highly-rated skills.
zgsm-ai/costrict
roo-translation
Provides comprehensive guidelines for translating and localizing CoStrict extension strings. Use when tasks involve i18n, translation, localization, adding new languages, or updating existing translation files. This skill covers both core extension (src/i18n/locales/) and WebView UI (webview-ui/src/i18n/locales/) localization.
zgsm-ai/costrict
roo-conflict-resolution
Provides comprehensive guidelines for resolving merge conflicts intelligently using git history and commit context. Use when tasks involve merge conflicts, rebasing, PR conflicts, or git conflict resolution. This skill analyzes commit messages, git blame, and code intent to make intelligent resolution decisions.
davila7/claude-code-templates
verl-rl-training
Provides guidance for training LLMs with reinforcement learning using verl (Volcano Engine RL). Use when implementing RLHF, GRPO, PPO, or other RL algorithms for LLM post-training at scale with flexible infrastructure backends.
davila7/claude-code-templates
openrlhf-training
High-performance RLHF framework with Ray+vLLM acceleration. Use for PPO, GRPO, RLOO, DPO training of large models (7B-70B+). Built on Ray, vLLM, ZeRO-3. 2× faster than DeepSpeedChat with distributed architecture and GPU resource sharing.
davila7/claude-code-templates
gguf-quantization
GGUF format and llama.cpp quantization for efficient CPU/GPU inference. Use when deploying models on consumer hardware, Apple Silicon, or when needing flexible quantization from 2-8 bit without GPU requirements.
davila7/claude-code-templates
Claude Code Guide
Master guide for using Claude Code effectively. Includes configuration templates, prompting strategies "Thinking" keywords, debugging techniques, and best practices for interacting with the agent.
Didn't find tool you were looking for?