ai Research Brief Generator

Turn 5 minutes of research questions into a 2,000-word structured brief with source scoring, bias assessment, and actionable recommendations.

The Problem

Your VP asks "what's the competitive landscape for X?" and you spend 3 hours cobbling together a Google Doc from 15 browser tabs, half of which are behind paywalls. The brief you produce is a wall of unstructured text with no source credibility scoring, no bias flags, and no recommendations — just "here's what I found." Decision-makers skim it, miss the key insight, and make the call based on gut instinct anyway.

What You Get

• Multi-source search strategy — executes primary, adjacent, and contrarian queries across 8-12 sources per brief to eliminate blind spots and capture dissenting perspectives
• Source credibility scoring — rates every source on a 0-10 credibility scale using peer-reviewed journals, established research firms (Gartner, McKinsey), reputable news outlets, and domain expert blogs
• Recency and bias assessment — flags each source as Neutral, Industry-favorable, Industry-critical, or Self-interested, with recency scores from 0 (3+ years old) to 10 (under 3 months)
• Structured executive brief output — produces an 8-section Markdown document: Executive Summary (150-250 words), Key Findings (5-8 numbered), Source Analysis table, Recommendations (3-5 with confidence labels), Risk Assessment matrix, Open Questions, and Methodology Note
• Consensus vs. contested claims — separates findings into consensus (3+ independent sources), emerging signals (1-2 sources), and contested claims (credible sources disagree), preventing misleading certainty
• Quality self-check — verifies every key finding has an inline citation, no single source dominates >50% of citations, and all recommendations are actionable (not "monitor the situation")

Why This Beats Prompting It Yourself

An unguided prompt like "research the market for X" produces an output that reads like a Wikipedia summary — no source hierarchy, no bias flags, no risk matrix, and no open questions. The agent will confidently present a single source's claim as consensus fact. This skill forces source triangulation (minimum 2 independent sources per key claim), separates signal from noise, and structures the output so a reader can scan the Executive Summary alone and walk away informed.

Example Session

You: Brief me on the state of the European electric vehicle charging market for 2026.

Agent: Executes 6 search queries across primary ("EU EV charging market 2026"), adjacent ("charge point operator consolidation Europe"), and contrarian ("EV charging market risks overcapacity"). Scores 11 sources (credibility 4-9, recency 3-8, bias flags: 2 Industry-favorable, 1 Self-interested). Produces 8-section brief with Executive Summary highlighting 3 consensus findings (market growing 34% CAGR, CPO consolidation accelerating, grid constraint emerging as bottleneck) and 2 contested claims (battery swap vs. fast charge). Recommends: "Invest in grid-connected CPOs" (High confidence), "Monitor battery-swap entrants" (Medium confidence).

Use Cases

• Preparing a competitive intelligence brief for a board meeting on 48-hour notice
• Synthesizing market research across 10+ sources before a product launch decision
• Creating a technology landscape report for a CTO evaluating 3 vendor options
• Producing a regulatory impact analysis from government publications and industry commentary
• Building a weekly intelligence digest on a specific market segment for an investment team

Known Limitations

The brief generator relies on web search and cannot access paywalled content (McKinsey full reports, Bloomberg Terminal data). Sources behind paywalls are noted as "[Paywalled — not fully verified]" in the source table, reducing the brief's authority on topics where the best analysis lives behind subscription walls.

Tags: research analysis briefings market-intelligence competitive-analysis

Version: 1.0.0