The Specter MCP connects your AI client straight to Specter's private-markets data, so your assistant can pull live company, people, investor and transaction records, and the signals that track how they change. Ask a question in plain English and it picks the right tool, resolves the entity, and returns the requested data.

This article lists every tool the MCP exposes and what each one does, then walks through a few common workflows, with a focus on talent and investor-interest signals. For setup, see the Specter MCP Installation Guide; for how sign-in works, see How Specter MCP Authentication Works.

How the tools are organized

The tools group by what you are working with. Most are read-only and return data. A handful are write tools that change your workspace (creating a list, saving a search); your AI client will ask you to confirm before it runs one.

Resolve an entity

Start here when you already know who you are looking for. Each tool turns an identifier you hold into a Specter record.

Tool	What it does
`find_company`	Resolve a company you can name (domain, LinkedIn, ID or exact name) to its minimal profile.
`find_person`	Resolve a known person (LinkedIn, email, ID or name) to its minimal profile.
`find_investor`	Resolve an investor firm (domain, name or ID) to a minimal profile.

Search and discovery

Use AI search tool to run a natural-language search across any product and get a compact result list

Tool	What it does
`search`	Run a natural-language search across any product and get a compact result list, plus a query you can save.

Company intelligence

Tool	What it does
`get_company_profile`	Pull a full company profile, presented as a mini investment memo.
`get_company_intelligence`	Get headcount by department, investor signals, founders and key people, plus the latest reported and estimated revenue and valuation.
`get_company_competitors`	See a ranked competitive landscape and the dimensions the market splits on.
`get_company_funding_rounds`	List the company's funding rounds with valuations and investors.
`get_company_acquisitions`	List acquisitions the company made or was part of.
`get_company_ipos`	List the company's IPO events.
`get_company_latest_revenue`	Get the most recent reported and estimated revenue, including high and low estimate ranges.
`get_company_latest_valuation`	Get the most recent reported and estimated valuation, including high and low estimate ranges.
`get_company_news_signals`	Pull news-sourced signals for the company: funding, revenue, traction and deals.
`get_company_investor_interest_signals`	See the investor-interest signals attributed to one company.

People

Tool	What it does
`get_person_profile`	Pull a full person profile with career timeline, education and highlights.
`get_person_email`	Get a personal or professional email for a person.
`get_person_talent_signals`	See the career moves Specter tracked for one person: founding, going stealth, exec hires, promotions.

Investors

Tool	What it does
`get_investor_profile`	Get an investor firm's firmographics and investment activity.
`get_investor_funding_rounds`	List the rounds an investor took part in, with the funded company and co-investors.
`get_investor_interest_signals`	See the interest signals sourced from one investor.

Lists

Tool	What it does
`get_lists`	List the company, people or investor lists you can access.
`get_list_items`	Pull the items in a specific list.
`create_list`	Create a new list (write)
`update_list`	Rename a list, change whether it is shared with the API, or add and remove its items. (write)
`delete_list`	Archive a list you manage. (write)

Record memberships and CRM context

get_entity_memberships checks which accessible Specter Lists contain a record and what your connected CRM holds on it. It supports up to 100 records in one call.

The response contains:

specter_lists: Lists you own, Lists shared with you, and organisation-public Lists that contain the record.
crm_records: CRM list memberships with per-entry status, record owner, lifecycle stage, linked deals, last and next activity, and sync time, where available.

Use it to answer “Which lists is this company in?” or “What does our CRM already hold on these records?”

Saved searches

Note: create_saved_search and delete_saved_search are write tools and will ask you to confirm.

Tool	What it does
`get_saved_searches`	List your saved searches across any product. Each search includes `last_viewed_at`: when you personally last opened it, rather than when a colleague did.
`get_saved_search_items`	Pull results from a saved search: the complete set, additions since a date you choose, or additions since you last opened it using `since_last_viewed=true`. In last-viewed mode, Specter resolves your personal last-visit time and applies the same cutoff to both the count and the returned page.
`create_saved_search`	Save a search you have built so it persists and is shared with the API. (write)
`delete_saved_search`	Delete a saved search you own. (write)

When requesting additions, the response includes applied_cutoff and cutoff_source, so your AI client can state the window covered.

With since_last_viewed=true, you don’t need to supply a date. Saved Search additions are tracked for 60 days:

If you’ve never opened the search, results cover the available tracking window and cutoff_source is never_viewed.
If your last visit is older than the tracking window, results cover the available tracking window and cutoff_source is retention_limit.

Common workflows

You do not call these tools by name. You describe what you want, and your assistant chains the right tools together. The usual path runs from broad to specific:

1. Search in plain English

Ask something like "Find high-growth fintech companies in Europe" and your assistant uses search to return matching companies, people or investors. Describe what you are after in plain terms.

2. Save a search to track it

Ask your assistant to save a search with create_saved_search. When you return, ask “What’s been added since I last opened this saved search?” Your assistant uses get_saved_search_items with since_last_viewed=true. Specter uses your own last-visit time, applies the same cutoff to the count and results, and returns the window used. If you’ve never opened the search, or your last visit was more than 60 days ago, it returns additions within the available tracking window and identifies the fallback.

3. Build a target list

Ask "Create a list called Q3 Targets with these companies" and your assistant resolves each one with find_company, or pulls them straight from a search, then creates the list with create_list. Use update_list to add, remove or rename items later.

4. Go deeper on a company

Once you have a company in front of you, ask for more and your assistant follows up with the detail tools, for example get_company_profile for a full memo, get_company_competitors for the competitive set, get_company_funding_rounds for funding history and much more

5. Check what your team already holds

Ask “Which Specter Lists is this company in, and what does our connected CRM hold on it?” Your assistant uses get_entity_memberships to return accessible list memberships and available CRM context. You can check up to 100 records in one call.

Need a hand?

Open the Intercom chat in your Specter dashboard or email support@tryspecter.com. We are quick to respond and always keen to hear your feedback.

Specter MCP - Tools & Workflows