Briefs Finance Connector
Question 1 of 2

What do you want your portfolio to do?

Which sounds more like you?

You're a General Purpose investor

Your copy-paste prompt below is now personalized to this profile — it leads with the right Briefs Score archetype.

Watch: connecting Briefs Finance to Claude

Bring Briefs Finance into Claude or ChatGPT

Investment research, financial scoring, watchlists, and portfolio coaching — right inside the AI tools you already use, secured behind your Briefs account.

How to connect with other providers

Works with any connector-compatible client — pick yours below.

  1. 1

    Add the connector

    Use a one-click link, paste a command, or add https://mcp-dev.briefs.co/mcp manually in your client's connector settings.

  2. 2

    Sign in via OAuth

    Your client opens a browser window. Approve access with your existing Briefs account — no API keys to manage.

  3. 3

    Call tools

    Try whoami or help first to confirm the connection, then explore the catalog below.

Claude

Recommended

claude.ai on web, desktop, and mobile share one connector config.

Add to Claude

Opens Claude's Add custom connector dialog prefilled with this server — you review and confirm. Works on Free, Pro, and Max. Team and Enterprise admins: use Organization settings → Connectors with the same link.

Manual: Customize → Connectors → Add custom connector.

Claude Code

CLI coding agent.

claude mcp add --transport http briefs-finance https://mcp-dev.briefs.co/mcp

The OAuth flow opens automatically in your browser. Verify with claude mcp list.

Cursor

One-click install via deep link.

Add to Cursor

OAuth uses automatic dynamic client registration -- nothing to configure.

Manual: Settings → MCP → New MCP Server.

VS Code

GitHub Copilot agent mode.

Add to VS Code
code --add-mcp '{"name":"briefs-finance","type":"http","url":"https://mcp-dev.briefs.co/mcp"}'

ChatGPT

Requires Plus or Pro (or Business/Enterprise with admin enablement) — not available on ChatGPT Free.

  1. Settings → Security and Login → Developer Mode — toggle it on (the first toggle only) and exit
  2. Navigate to Plugins in the left sidebar
  3. Click + to add a connector — Name: "Briefs Finance", Connection: https://mcp-dev.briefs.co/mcp
  4. Click Create, then Add
  5. Complete the OAuth sign-in

Windsurf

Add to ~/.codeium/windsurf/mcp_config.json.

{"mcpServers":{"briefs-finance":{"serverUrl":"https://mcp-dev.briefs.co/mcp"}}}

If OAuth sign-in doesn't trigger, Windsurf may need the mcp-remote local proxy — see the Windsurf docs.

Gemini CLI

Add to ~/.gemini/settings.json.

{"mcpServers":{"briefs-finance":{"httpUrl":"https://mcp-dev.briefs.co/mcp"}}}

OAuth is discovered automatically on the first 401.

Manual configuration

Any other connector-compatible client: Streamable HTTP transport, OAuth 2.0 browser-flow auth, no API keys.

Connector Endpoint
https://mcp-dev.briefs.co/mcp
Connecting to remote MCP servers

Available Tools (63)

Grouped by capability, derived live from the registered tool catalog.

Get earnings transcript
tier: proscope: briefs:readwrites: no

Read a company's earnings call transcript. By default returns the latest earnings call: its title, call date, fiscal period (e.g. "Q3 FY2026"), participants with roles, and its first few sections, plus a list of the ticker's other calls (newest first, up to 20, with available_total saying how many exist). Each section is one whole speaker turn, labelled prepared_remarks or q_and_a, never cut mid-answer. To read on, call again with start_index set to the previous response's next_start_index (null means you reached the end). To read a different call, pass transcript_id from the available list, or fiscal_year with fiscal_quarter. A call listed as "Special call" is not a quarterly earnings call. One required argument: ticker (a real symbol, e.g. AAPL). Quote speakers by name and role, and say which call and fiscal period a quote comes from. Requires a Briefs Pro subscription. Side effects: None Idempotency: Safe to retry

Coach foundation build
tier: nullscope: briefs:readwrites: no

Coach the user through a financial-foundation framework — debt payoff, emergency savings, or planning — chosen by the `topic` param (debt | emergency_savings | planning | analyze). Reads the continuity spine (get_coach_context) plus optional capture-only financial-situation answers, then returns a meta-prompt that teaches the framework generically with illustrative example math (avalanche vs snowball, the 3-6 month emergency-fund rule, the foundation ordering). Educational and impersonal state only (BT-883) — never a buy/sell/hold recommendation, never a personalized 'you should' directive, and never the user's captured figures in math. Capture happens via set_profile_answer (pointed to by the prompt) — this tool writes nothing and never moves risk level or guardrails (those are set by profile answers). Personalized figures are Ring 3 (Plaid/BT-440). May reference Briefs content (search_content) when available. Requires briefs:read scope. Side effects: None Idempotency: Safe to retry

Thesis discover
tier: nullscope: briefs:readwrites: no

Start a Thesis Finder discovery session from a free-text investment idea (seed). Returns one of two shapes keyed on `kind`: (1) a **Question** (`kind: "question"`) when the seed is ambiguous — it carries a `prompt`, selectable `options` (each with an `option_id`), and an opaque `state` blob. Present the options to the user; when they pick one, call **thesis_refine** with that `option_id` and the `state` blob echoed verbatim from this tool's resource. The option id "__other__" means the user supplied free text — pass it as refine's free_text. (2) a **DiscoveryResult** (`kind: "result"`) with matched ETFs, a thesis summary, and sources — skip refine and call **thesis_plan** with the `result` blob echoed verbatim from this tool's resource. If the user has exhausted their monthly discovery limit, returns a readable upgrade prompt instead of a result. Requires briefs:read scope. Side effects: Consumes one Thesis Finder query against the monthly cap (Plus tier; Investor, Pro, and free are uncapped) Idempotency: Not idempotent (each call consumes a query and may start a new session)

Get briefs score
tier: investorscope: briefs:readwrites: no

Return the Briefs investor-type scores for a stock ticker — one per investor archetype (Income, Wealth Preservation, Growth). Each archetype returns a letter grade and a neutral-labelled 7-axis snowflake (axis label → 0–100 value) plus a status (scored / na). The scores are deterministic and educational; the underlying methodology (weights, formulas, thresholds) is not exposed. Call get_score_glossary to learn what each axis label means. Returns score_available=false when no current Briefs score can be served for the symbol — check unavailable_reason to tell why, and whether waiting/retrying will ever help: "not_yet_scored" means a real ticker whose score simply hasn't been computed/frozen yet (safe to retry shortly); "ticker_not_found" means the symbol itself could not be resolved by any data vendor — e.g. a company name ("Apple") or a misspelled ticker ("APPL") instead of the real symbol ("AAPL") — retrying won't help, the caller should re-check the symbol instead; "not_trading" means the instrument is no longer actively trading (delisted or long-suspended) — terminal: retrying will not help, and this response carries no grades even though an old grade may once have existed — do NOT present the symbol as a current holding or a current grade; tell the user the stock is not trading; "store_unavailable" means an infra/store outage, also safe to retry. This tool does NOT resolve a company name to a ticker — pass an actual ticker symbol. Requires a Briefs Investor subscription. A “Briefs score” is Briefs’ proprietary, named investor-type score — not a generic financial score. The briefs_score in this tool’s name is intentional and refers to that named concept; see get_guidelines for what a Briefs score is and how the score tools fit together. Side effects: None Idempotency: Safe to retry

Get profile answers
tier: nullscope: briefs:readwrites: no

Get the current user's answers to all profile questions. Returns the question id, label, input type, and the user's answer (text or selected options). Use this to see what the user has already answered. Side effects: None Idempotency: Safe to retry

Create watchlist
tier: nullscope: briefs:writewrites: yes

Create a new named watchlist and return its id. Use this when the user wants a brand-new list/sleeve (e.g. "AI Semis") — not to add a stock to a list they already have (that is edit_watchlist_items' add entries). Takes a name plus an optional description, and the return value carries the new numeric id that sibling tools expect as watchlist_id / target_watchlist_id — take it from here, or re-read it later with list_watchlists. Optionally pass initial_tickers (an entry per stock: ticker + exchange required, company_name / notes / target_price / weight / tags optional) to seed the list in the same call — building a whole list is then one call instead of an add per ticker; each seed is applied one at a time and a seed that fails (e.g. a duplicate ticker) is reported against that entry only, without rolling back the ones before it. A duplicate name returns a conflict rather than a second list. Requires briefs:write scope. Side effects: Creates a watchlist row; with initial_tickers, also creates one item row per seed Idempotency: Safe to retry (duplicate name returns conflict, no duplicate created)

Create bookmark
tier: nullscope: briefs:writewrites: yes

Save a content item as a bookmark. Supports brief, article, lesson, and post types. If no collection_id is specified, the item is saved to the default 'Saved' collection. Requires briefs:write scope. Side effects: Creates a bookmark row (in the default 'Saved' collection if none specified) Idempotency: Safe to retry (duplicate content returns conflict, no duplicate created)

Create comment
tier: proscope: briefs:writewrites: yes

Create a comment on a community feed post. Provide the post_id and text content. Posts containing external URLs will be auto-flagged for review. Requires a Briefs Pro subscription. Admins and support staff with community:moderate scope act with elevated privileges. Call get_post_thread first to get the post_id and understand the conversation context. Side effects: Creates a comment row; may auto-flag for moderation Idempotency: Not idempotent (each call creates a new comment)

Get article
tier: nullscope: briefs:readwrites: no

Read a single published Briefs news article by id (from search_articles). Returns the title, link, publish date, and full article content as HTML. Only published articles are available; drafts are not returned to members. Side effects: None Idempotency: Safe to retry

List reports
tier: nullscope: briefs:readwrites: no

List INDIVIDUAL Briefs reports, newest first (content-identity design: a report lesson like "Briefs Pro Reports | 2026" is a year-archive, not a report itself — each entry here is one specific report, e.g. "Growth Report | August 2026"). Each entry has the content_id (the primary key for read_report_section), the parent lesson_id/lesson_slug and module_slug, title, published_on, minimum tier, whether it's locked for your current subscription tier, and has_text (false for a stub entry with no extracted body text — skip those). The FIRST entry is the latest report. collection scopes the listing (default research = Briefs research/Pro Reports/newsletter reports, unchanged behavior — pass portfolio or all to reach Briefs Fund portfolio reports as well; each portfolio report is its own entry with its own content_id, flagged locked above your tier). Use search_report_content to find a specific passage across reports by natural language, or read_report_section with the content_id to page through a specific report deterministically (works even for a just-published report search hasn't ranked yet). Side effects: None Idempotency: Safe to retry

Create chat project
tier: nullscope: briefs:writewrites: yes

Create a new chat project (folder) for organizing chat sessions. Requires briefs:write scope. Side effects: Creates a chat project (folder) row Idempotency: Not idempotent (each call creates a new project)

Skills

Prebuilt skills for common Briefs workflows are coming. Tell us which workflows you'd like us to build first.

Tell us what you'd like

Frequently asked questions

How do I set up the Briefs Connector?

See Connect above for step-by-step setup instructions for your AI client. If you need more assistance, email us at hello@briefs.co, or call us at 1-888-259-5651.

What is the Briefs Connector?

The Briefs Connector (currently in Beta) links your Briefs Finance account to an AI assistant you already use (for example: Claude, ChatGPT, or Perplexity) so you can ask questions about Briefs research in plain language and get answers that draw on research and information available through your Briefs account.

It can access your Briefs Pro Reports and the Briefs Score to help your AI respond to your questions. The AI assistant, not Briefs, generates the response you see.

Do I need to be a Pro member?

Yes. Currently, only Briefs Pro members can access it.

What can the Briefs Connector do?

It can look up the Briefs Score for a stock and explain what each part of the score measures.

It can help you screen for stocks that hit a minimum grade. For example, you could ask “Give me a list of 10 growth stocks with a Briefs Score of B+ or higher.”

You can search across Briefs Pro Reports for what we said about a topic, and dig deeper on a topic or ticker that interests you.

The Connector is designed to help you research and evaluate investments using Briefs information. It does not place trades, access your brokerage account, or make investment decisions for you.

How much does it cost?

As of September 2026, the Briefs Connector is included with your Briefs Pro membership.

Note: Your AI provider may charge you separately for their plan. Refer to your AI provider's website for details.

Which AI assistants does the Briefs Connector work with?

As of September 2026, the Briefs Connector works with supported Claude accounts, including certain free and paid plans. Usage limits and available features may vary by plan and can change at any time.

For ChatGPT, availability may depend on your ChatGPT plan and settings.

Answers will vary depending on which assistant you use, which model it's running, and how you phrase your question. AI provider features, requirements, and limits can also change, so check your AI provider's terms and current product information for details.

What can I ask it?

Here are some starting points:

“What's the Briefs Score for [enter a stock ticker], and what's driving it?”
“Show me growth stocks with a Briefs Score of B+ or better.”
“What did the August Briefs Pro Report say about income stocks?”
“Explain what the moat is in the Briefs Score.”

Why is my AI saying there is a suspicious prompt?

Your AI may occasionally display a security warning when connecting to or using third-party tools. If prompted, review the information your AI provider presents and confirm that you intended to connect to Briefs Finance.

Will it help me decide what to buy or sell?

It can help you research and evaluate potential investments, but the investment decisions are yours. The Briefs Connector is not designed to tell you what you personally should buy, sell, or hold. Briefs Finance is an investing education and research platform. The Briefs Connector works by making Briefs research, data, and investment-rating methodologies available through the AI assistant you choose. The Connector is designed to make that research easier to explore so you can reach your own conclusions.

Briefs does not control how a third-party AI interprets or presents that information, and AI-generated responses may be incomplete or inaccurate. Before making an investment decision, consider reviewing the underlying Briefs research and other information you believe is relevant. Briefs does not provide individualized investment advice or do so through the Connector.

What is the Briefs Score it keeps referencing?

The Briefs Score is our proprietary scoring system, and provides three easy-to-read letter grades focused on different investment objectives.

Our three score types are: Growth, Income, and Wealth Preservation.

They are calculated using financial data from sources that include U.S. exchanges and SEC filings, together with our proprietary calculations and subscores.

Our scores are back-tested against up to 30 years of financial data. The formula and methodology is verified by our experts every week.

Back-tested results are hypothetical. They show how the scoring methodology would have performed using historical data and do not represent actual investment results. Past or back-tested results do not guarantee future performance. The Briefs Score is a research tool, not a prediction of how a security will perform.

Why does one company have three different grades?

Because a stock can score differently depending on the investment objective being evaluated. Every stock gets a separate grade for growth, income, and wealth preservation. An A for income and an A for growth measure different things and don't mean the same thing.

If you want to learn more about these different investing approaches, check out our articles on the topic: Growth Investing, Income Investing, and Wealth Preservation Investing.

Why can't it find a score for my stock?

There are rare occasions where a stock isn't scored.

Data for that security may be missing or incomplete; this is often resolved in a matter of minutes.

Securities not listed on U.S. exchanges, or ETFs that have a large number of unlisted securities, may also lack scores.

Other times it can just be a tech issue. You may want to try some common troubleshooting steps: log out and back into your AI, or try to disconnect and reconnect the Briefs Connector.

If you need assistance, email us at hello@briefs.co, or call us at 1-888-259-5651.

Can my AI get things wrong?

Any AI can take information, such as from the Briefs Connector, and present it in different ways, or sometimes generate information that is inaccurate or unsupported (this is sometimes called an “AI hallucination”). Remember, AI assistants can make mistakes.

The Briefs Connector gives your AI access to Briefs information, but it cannot guarantee that the AI will interpret, summarize, or present that information correctly. For important information, you can check the underlying Briefs Pro Report or Briefs Score directly.

Note: You can always log in to finance.briefs.co to access your Briefs Pro Reports or check the Briefs Score on the web if you want.

What does the Briefs Connector do with my data?

When you connect the Briefs Connector, you're giving your AI provider permission to request information from your Briefs Pro account on your behalf. The Connector only provides information your authenticated Briefs account is authorized to access.

Information you submit to your AI provider is handled under that provider's applicable terms and privacy practices. Briefs separately handles information it receives or processes through the Connector in accordance with its own applicable privacy terms. We recommend reviewing your AI provider's privacy and data-use settings so you understand how it may use or retain information you share with it.

Can my AI provider see my Briefs data?

Yes. Information retrieved through the Briefs Connector must be provided to your AI provider so that it can generate a response. How your AI provider processes, uses, stores, or retains that information is governed by its applicable terms, privacy policy, and your account settings.

Connecting the Briefs Connector therefore means that information returned from your Briefs account may be shared with your selected AI provider. Review that provider's privacy and data-use settings if you have questions about how it handles that information.

How do I disconnect it?

To remove the Briefs Connector, you'll need to find it in your AI's connector settings and disconnect it there. Nothing in your Briefs Pro account changes.

Disconnecting the Connector stops future access through that connection, but it may not delete information previously provided to or retained by your AI provider. Check your AI provider's settings and policies for information about deleting prior chats or data.

I still need help. Who do I ask?

If you need assistance, email us at hello@briefs.co, or call us at 1-888-259-5651.