More screenshots

    Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIDeepSeekDeepSeekMuseMuse+22 more

    Browser Research to Cited Report

    1

    Turn a research question into a decision-ready report with source grades, claim-to-source mapping, contradictions, and confidence levels.

    $19

    Secure checkout via Stripe

    Money-back guarantee

    Converts to your local currency at checkout

    Security scanned
    Browser Research to Cited Report

    Browser Research to Cited Report

    Example session with this skill installed

    Research question

    Should our 20-person B2B SaaS company add an AI meeting-notes feature to our existing project-management product?

    Context

    • We sell to operations managers at mid-market companies.
    • Our current price is about $50 per user per month.
    • We do not have an in-house machine-learning team.
    • We can use third-party transcription and summarization APIs.
    • We want to know whether this feature is a differentiator or becoming table stakes.
    • We need evidence on buyer demand, competitor presence, pricing impact, and implementation risk.
    • The decision is whether to build a prototype next quarter.

    Constraints

    • Limited budget.
    • No desire to build custom speech models.
    • Must avoid heavy compliance burden.
    • Preference for sources from the last 24 months.

    Output needed

    • A cited research report.
    • Evidence table.
    • Source quality grading.
    • Contradictions.
    • Recommended next action.
    • Read your context and instructions
    • Compiled the browser research to

    Cited Research Report

    1. Research Brief

    | Field | Value |
    | Research question | Should a 20-person B2B SaaS company add AI meeting notes to its project-management product
    | Decision supported | Whether to build a prototype next quarter. |
    | Overall confidence | Medium |
    | Decision readiness | Conditional |

    2. Executive Summary

    Direct answer: Build a narrow prototype only if it integrates into existing project workflows.
    Standalone AI meeting notes appear likely to become table stakes; differentiate via meeting → tasks, owners, deadlines, and project updates.

    3. Key Findings

    1. Basic AI meeting summarization is becoming common. | Medium | S1, S2
    2. Stronger differentiator is converting notes into tasks and project updates. | Medium | S3
    3. Implementation risk is moderate (third-party APIs + privacy). | Medium | S4

    5. Source Register

    S1 Competitor A pricing | vendor | Tier A | vendor-self
    S2 Competitor B AI features | vendor | Tier C | vendor-self
    S3 Customer interview summary | user-provided | Tier B | user-provided
    S4 Transcription API guidance | secondary | Tier B | independent

    6. Contradictions

    Table stakes (S1,S2) vs workflow-native differentiator (S3) → prototype the workflow layer.

    9. Recommended Next Actions

    • Interview 8–12 customers about meeting-to-task needs
    • Compare three transcription providers on privacy, pricing, reliability
    • Prototype only the meeting-to-task flow if interviews are positive

    10. Research JSON

    { "overall_confidence": "Medium", "decision_readiness": "Conditional", ... }

    Note: Demo sources are illustrative placeholders. Real runs use verified accessible sources only.

    Connects securely to your tools. The creator never sees your data.

    About this skill

    The problem

    Generic “research this” prompts produce summaries without defensibility: missing sources, vendor marketing treated as fact, contradictions buried, and no clear signal whether evidence is strong enough to decide. Teams waste hours rebuilding the same structure in Docs or Notion.

    What it does

    Browser Research to Cited Report runs a disciplined research workflow and returns a decision-ready artifact:

    • Scoped research brief (question, decision, includes/excludes, timeframe)
    • Executive summary with a direct answer, implications, and key unknowns
    • Key findings with confidence levels (High / Medium / Low / Insufficient)
    • Evidence table mapping every material claim to source IDs
    • Source register with quality tiers (A–D) and independence labels
    • Contradictions, assumptions, missing data, and “what would change this conclusion”
    • Recommended next actions
    • Structured Research JSON for Notion, Airtable, CRM, or internal tools

    Frameworks & tools

    • Source quality rubric (Tier A–D) and independence rules
    • Research integrity rules (no fabricated URLs/quotes/stats; vendor claims labeled)
    • Report schema for Markdown + JSON parity
    • Agent web search / browse / fetch tools when available
    • Safe degradation when only user-provided excerpts are available

    Why this beats prompting it yourself

    • Enforced claim-to-source mapping — no orphan conclusions
    • Source grading and independence so marketing pages don’t masquerade as proof
    • Contradictions and confidence levels instead of false certainty
    • Decision readiness (Sufficient / Conditional / Insufficient) built into the brief
    • JSON artifact ready to reuse — not a one-off chat dump
    • Clear safety boundaries: public sources only; no paywall bypass, send, or publish

    Use cases

    • Should we ship this feature next quarter?
    • Competitor landscape with evidence, not vibes
    • Vendor comparison before a contract
    • Client or internal due-diligence memo
    • Market or trend research with recency constraints
    • Leadership brief that separates fact, estimate, and interpretation

    Known limitations

    • Quality depends on accessible public sources (or user-provided excerpts)
    • Cannot bypass paywalls, logins, CAPTCHAs, or private systems
    • Not legal, medical, tax, or financial advice — research findings only
    • Fast-moving topics need recent sources; older material is background only
    • If browser tools are unavailable, output is limited to user-provided material or a research plan
    • Sample sources in the marketplace demo are illustrative placeholders, not live citations

    Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Skills for Research & Analysis.

    • Research quality depends on accessible public sources or user-provided excerpts.
    • Does not bypass paywalls, logins, CAPTCHAs, robots.txt, or private systems.
    • Not legal, medical, tax, or financial advice — findings only; recommend qualified review in regulated domains.
    • Fast-moving markets require recent sources; older sources are background.
    • Without browser/search tools, the skill uses only user-provided material or outputs a research plan.
    • Marketplace demo sources are fictional/illustrative placeholders labeled as such.

    How to use

    1. 1

      Connect your agent to the Agensi MCP

      Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup

    2. 2

      Sign in to Agensi

      Your agent shows a sign-in link. Sign in and click Allow.

    3. 3

      Ask for what you need

      Your agent loads this skill and uses it on your task. It always asks before buying a skill.

    Reviews

    No reviews yet

    Be one of the first to try it. Every listed skill goes through the trust checks below.

    Security scanned

    Checked by our 8-point scan before listing

    Fresh listing

    Recently published to Agensi

    Money-back guarantee

    Doesn't do what its page promises? You get your money back.

    Trust & safety

    Security scan

    Security scan passed 20 days ago

    Permissions

    Read Files
    Browser
    Network Access

    File Scopes

    User-provided
    paths
    only
    (
    if
    any)

    Read-only research by default. Browser/Network used only for public pages. Does not log in, bypass paywalls, CAPTCHAs, or robots.txt. Does not send email, publish, contact sources, or exfiltrate data. If auto-detect grants Shell/Write/Env, turn them off.

    Listed3 months ago

    What's inside

    Frequently Asked Questions

    Unlike standard chat prompts that summarize text, this skill enforces a disciplined rubric including source grading (Tier A-D), explicit claim-to-source mapping, and confidence level tags to ensure your reports are defensible and decision-ready.

    The skill is designed to work with any agent capable of web browsing or document fetching; if tools are unavailable, it can still process provided text excerpts into the structured report format.

    The primary output is a Markdown-formatted report for human reading, accompanied by a structured JSON artifact that allows you to instantly sync research data into Notion, Airtable, or your CRM.

    The report includes a dedicated section for 'Contradictions and Missing Data' and assigns confidence levels (High to Insufficient) to findings, signaling exactly where evidence is weak or conflicting.

    You receive the full research workflow configuration, the specialized report schema, and the source-grading rubric, allowing you to deploy the skill immediately within your supported AI environment.

    The skill is restricted to public web sources and cannot bypass paywalls, logins, or CAPTCHAs, ensuring all research is gathered through ethical and standard browsing protocols.

    New in Research & Analysis