Cookbook: Research Agent

This example combines:

• loop reasoning
• built-in tools
• semantic memory
• eval-based quality checks

1. Goal

Build a research assistant that can:

1. use tools for retrieval and processing
2. store distilled findings as semantic memory
3. evaluate output quality before release

2. Skeleton

import { HarnessBuilder } from 'colony-harness'
import { createBuiltinTools } from '@colony-harness/tools-builtin'
import { runEvalSuite, containsScorer } from '@colony-harness/evals'

// Build harness, run task, then evaluate output quality.

3. Design notes

• restrict tool permissions (run_command allowlist)
• store reusable conclusions, not noisy raw logs
• run evals in CI before release

4. Minimum pre-release checks

1. pnpm build
2. pnpm typecheck
3. pnpm test
4. pnpm docs:build
5. pass-rate threshold for key eval cases