# Dealroom MCP — Tools & Filters Reference

A machine-readable reference for the Dealroom Model Context Protocol (MCP) connector.
Point your AI assistant at this file (or paste it in) so it knows exactly which tools
and filters are available and how to use them.

- **Endpoint:** `https://mcp.dealroom.co/mcp`
- **Auth:** OAuth 2.1 sign-in with your Dealroom account. No token to paste.
- **Access:** Free with any Dealroom Premium seat (for a limited time).
- **Scope:** Read-only. Every tool reads Dealroom data; nothing is written back to your account.
- **Coverage:** Companies, investors, founders, universities, funding rounds and exits.

When prompting, name the connector explicitly — e.g. *"using the Dealroom connector"* —
so the assistant routes the query to Dealroom rather than answering from memory.

---

## Tools (22)

### Search — find entities
| Tool | What it does |
| --- | --- |
| `search_companies` | Find companies by keyword or concept, sector, stage and location |
| `search_investors` | Find investors by focus, thesis and geography |
| `search_universities` | Find universities and research institutions |
| `search_transactions` | Find funding rounds by type, size, date and place |

### Analyze — profile a single entity
| Tool | What it does |
| --- | --- |
| `analyze_company` | Full company profile: funding, investors, team, growth signals |
| `analyze_investor` | Investor profile: focus, portfolio, exits, funds, co-investors |
| `analyze_founder` | Founder track record and prior companies |
| `analyze_university` | University ecosystem and spinouts |
| `compare_companies` | Compare companies side by side |
| `similar_companies` | Find competitors and lookalikes |

### Entity — drill into one resolved entity (company or investor)
| Tool | What it does |
| --- | --- |
| `entity_fundings` | Funding rounds for the entity |
| `entity_investments` | Investments the entity has made |
| `entity_exits` | Exits and acquisitions |
| `entity_funds` | Funds raised or managed |
| `entity_analytics` | Growth signals: hiring, web traffic, app downloads |
| `entity_co_investors` | Frequent co-investors and shared deals |
| `entity_shareholders` | Shareholders and their stakes |
| `entity_lp_investments` | LP positions (funds the entity backs as a limited partner) |
| `entity_alumni` | Notable alumni and founders |
| `entity_events` | Key events and milestones |
| `entity_patents` | Patents |
| `entity_nearby` | Nearby companies by location |

---

## Searching & filters

Each search resolves a list you can hand to `analyze_*` or `entity_*` tools by `uuid`.

**Keyword vs. semantic** — `search_companies` / `search_investors` take a `keyword_type`:
- `default_next` (default): match by a known name (e.g. a specific company or fund).
- `semantic`: match by a concept, sector or thesis (e.g. `climate tech`, `b2b saas`).

**Location** — pass a place name (`Europe`, `Germany`, `Berlin`) plus a `location_type`
that controls which location field is matched (founding location, HQ, all locations,
regions, founders' universities, etc.).

**Structured filters** — richer constraints (sector, stage, funding range, founding year,
and more) go in `form_data` as `must` / `should` / `must_not` clauses, with optional
`currency` and AND/OR `execution`. The connector returns the **available filter keys and
their valid values at the bottom of every search response** — always use those exact keys
and values rather than guessing.

**Sort** (investors) — `investor_rank_total`, `investments_num`, `rounds_count_all_time`,
`rounds_count_12_months`, `investor_exits_num`, `investor_total_funding`, or `name`.
Prefix with `-` for descending (e.g. `-investor_total_funding`).

**Pagination** — `limit` (max 25) and `offset`.

---

## Usage tips

- **Resolve once, then drill in.** Resolve an entity with a `search_*` or `analyze_*`
  tool, then reuse its `uuid` across the `entity_*` tools — it's faster and unambiguous.
- **Chain in one prompt.** Find → enrich → act (e.g. add results to your CRM) in a single
  request. The connector is read-only, so the "act" step happens in your other tools.
- **Currencies are reported as raised.** Rounds come back in their original currency
  (e.g. a round may be in EUR while others are in USD) — they are not normalised.
- **Data is traceable.** Answers are live Dealroom records attributable to source, not the
  model's guesses.

---

*Generated for the Dealroom MCP connector. See [dealroom.co/mcp](https://dealroom.co/mcp/).*