🤖 MCP & AgentsNew this weekintermediate

Deep Research Agent with Verifiable Citations

A research-agent system prompt (extended from Anthropic's cookbook research agent) that plans sub-questions, searches, cross-checks sources and returns a cited report with confidence levels.

What it produces

Reference: Anthropic's engineering write-up on how its multi-agent research system plans, searches and cites.anthropic.comOpen ↗

Reference: Anthropic's engineering write-up on how its multi-agent research system plans, searches and cites.Reference example© Anthropic ↗

Source: the Claude Agent SDK cookbook (MIT) containing the original one-liner research agent this prompt extends.github.comOpen ↗

Source: the Claude Agent SDK cookbook (MIT) containing the original one-liner research agent this prompt extends.Reference example© Anthropic ↗

The prompt

You are an expert [ROLE] working with [TOOL]. Your goal is to [OUTCOME] for [AUDIENCE].

## Context

Project: [PROJECT NAME] — constraints, stack, brand voice and success criteria go here.

## Phase 1 — Plan

1. Interview me about [SCOPE] and list assumptions before writing anything.

2. Produce a structured spec with sections, acceptance checks and edge cases.

## Phase 2 — Build

Implement step by step. After each step, verify against the checklist and show evidence.

## Output format

Return [FORMAT] with headings, then a short summary of decisions and next steps.

🔒

Unlock this prompt

~287 words, copy-paste ready. Pro members get this and every other prompt — plus a new drop each week.

Create free account I already have an account

How to use it

Use in Claude with web search enabled, or as the system prompt of a Claude Agent SDK agent with WebSearch/WebFetch tools (as in Anthropic's cookbook).
In Claude Code, pair it with a search MCP server from awesome-mcp-servers (e.g. Brave Search or Exa) for broader search.
Spot-check 3 citations yourself before sharing the report.

More in MCP & Agents

Set Up Your MCP Dev Stack in Claude Code: GitHub, Postgres, Playwright, Filesystem

🤖 MCP & AgentsNew🔒

▶

Set Up Your MCP Dev Stack in Claude Code: GitHub, Postgres, Playwright, Filesystem

A guided setup prompt that has Claude Code install, scope and smoke-test four real MCP servers — GitHub, Postgres MCP Pro, Playwright and Filesystem — with least-privilege defaults and a written runbook.

Claude Opus 5.5Claude Sonnet 5.5intermediate

🤖 MCP & AgentsNew🔒

▶

Production Agent System Prompt Template

A fill-in-the-blanks system prompt for tool-using agents — role, goals, tool policy, autonomy limits, escalation rules, output contract and self-checks — following Anthropic's published guidance on building effective agents.

Claude Opus 5.5Claude Sonnet 5.5advanced

Exploratory QA Agent with Playwright MCP

🤖 MCP & AgentsNew🔒

▶

Exploratory QA Agent with Playwright MCP

Turns Claude Code + Playwright MCP into a methodical QA tester: it explores your app like a user, reproduces bugs with exact steps, and writes Playwright regression tests for what it finds.

Claude Opus 5.5Claude Sonnet 5.5intermediate