Quick Verdict
LittleSis MCP is a community TypeScript adapter by code-rabi that gives an MCP client read-only access to the public LittleSis database of people, organizations, and relationships. Its tools cover entity search, entity details, classifications, relationships, connections, lists, and individual relationship records. The current published package is [email protected]. It can be useful for investigative journalists, researchers, and civic-technology developers who want faster lead discovery. It must not be treated as an automated fact arbiter. The repository is tiny, with only a few stars and a small commit history, so maintenance continuity and security response are uncertain. More importantly, records about political donations, board service, employment, and social connections are easy for a language model to flatten into unsupported allegations. Every consequential output should be checked against entity identity, relationship type, dates, and cited source material.
Best For
- Investigative reporters and academic researchers exploring links among public figures and organizations.
- Developers building a read-only research assistant with visible citations and mandatory human review.
- Teams using Claude or ChatGPT to organize questions without delegating conclusions to the model.
- Civic organizations familiar with the US political and corporate-data context and prepared to handle corrections and attribution.
It is not suitable as the sole source for automated due diligence, risk scoring, adverse-action decisions, or public accusations.
Key Features
The adapter exposes get_entity, get_entities for batches up to 300, filtered search_entities, get_entity_extensions, get_entity_relationships, get_entity_connections, get_entity_lists, and get_relationship. These are retrieval tools. They do not edit the LittleSis database, publish claims, or take actions on behalf of a user.
The correct unscoped npm package is littlesis-mcp, and version 1.0.3 is the release to pin as of July 21, 2026. The repository README also documents running a built Node entry point from source. Before deployment, verify the npm publisher and integrity, inspect the lockfile and transitive dependencies, and confirm that the actual package entry point matches the client configuration. Older summaries that cite a scoped package should not be followed.
Use Cases
A careful investigation can begin by searching a name, resolving same-name entities, reading classifications, and then retrieving relationships around board membership, donations, employment, or organizational links. Researchers can organize several entities into a lead table and return to LittleSis pages and underlying public records for verification.
The model must preserve the difference between a recorded connection and wrongdoing. A shared board, donation, employer, or address does not establish coordination, intent, or misconduct. Before publication, retain the LittleSis entity and relationship IDs, relationship category, start and end dates, original source, retrieval date, and a note explaining any uncertainty. For a living person, editorial and legal review should occur before a generated narrative leaves the research team.
Pricing
The adapter is free and its code is licensed under MIT. Data obtained from LittleSis is offered under CC BY-SA 4.0, which requires attribution and may impose share-alike obligations on an adapted database. These licenses apply to different layers: MIT covers the adapter code; it does not convert the returned data to MIT.
Version 1.0.3 was released on August 5, 2025. By July 21, 2026, the project remains a small community effort rather than a staffed service. Pin the exact package version and integrity hash, retain a dependency inventory, and have a path to the upstream LittleSis API if the adapter stops working.
Pros
- Clear read-only tool surface for major LittleSis entity and relationship queries.
- No write operations that could accidentally modify upstream public data.
- Simple Node and MCP integration for a research prototype.
- Explicit separation is possible between the adapter’s MIT license and the data’s CC BY-SA terms.
- Useful as a retrieval layer alongside research orchestration concepts from LangChain.
Cons
- A two-star, roughly 13-commit project has limited maintenance and security-response capacity.
- The 1.0.3 release may age as MCP clients and dependencies change.
- LittleSis records can be incomplete, outdated, ambiguous, or missing context.
- Hallucinated connections or overconfident summaries can create defamation and reputational harm.
- Public-source aggregation can increase privacy impact, especially for non-public people.
- CC BY-SA data obligations are easy to overlook when the adapter itself is MIT.
Alternatives
| Alternative | Best when | Main difference |
|---|---|---|
| LittleSis website and API | Verifying records directly | Removes one model interpretation layer and shows source context more clearly |
| OpenSecrets | Researching US campaign finance and lobbying | More specialized around money in politics |
| SEC EDGAR | Checking corporate filings and officers | Authoritative primary filings, but relationship extraction takes more work |
| Wikidata SPARQL | Querying a broad open knowledge graph | Wider scope with different depth and sourcing for political relationships |
| Aleph | Linking investigative datasets and documents | A fuller investigation platform with heavier deployment and governance needs |
| Hugging Face | Building custom entity-extraction models | Provides models and datasets, not LittleSis relationship facts |
For high-risk investigations, a direct API or primary record should remain the evidence layer. MCP is only a convenience layer for discovery.
FAQ
What npm package and version should I use?
The current package specified for this page is [email protected]. Pin it and verify package integrity before installation.
Can the server modify LittleSis records?
The documented tools retrieve entities and relationships; they do not provide write operations.
Does the MIT license cover returned data?
No. MIT applies to the adapter. LittleSis data is provided under CC BY-SA 4.0.
Can model output be published directly?
No. Resolve identities, inspect relationship definitions and dates, follow source citations, and complete editorial and legal review.
Does public data eliminate privacy risk?
No. Aggregation, amplification, and mistaken identity can still harm people, particularly private individuals.
Is it appropriate for unattended enterprise due diligence?
No. Its small maintenance base, source coverage, and interpretation risks make it a lead-generation aid, not a sole due-diligence authority.
Bottom Line
LittleSis MCP is a thin and convenient bridge from public power-network data to an MCP client. Its responsible use is equally narrow: pin 1.0.3, distinguish MIT code from CC BY-SA data, use results only to generate leads, and trace every claim about a real person back to dated source material. Those controls matter more than the ease of adding the server to a client.