# plures-mcp [Health: Active]

**Category:** 💻 Developer Tools  
**Repository:** https://github.com/plureslabs/plures  
**GitHub Stars:** 0  
**npm Downloads (last month):** 2519  
**Views:** 0  
**Installs:** 0  
**Upvotes:** 0  
**Directory Page:** https://allmcps.com/mcp/plures-mcp

## Description
MCP server for Plures: lets any AI agent ask ID verified people to check a decision before it acts.

## Tools
Capabilities this server exposes over MCP:

- **ask_humans** — Send a decision to ID verified people before you act on it. Several people answer independently, each confirmed with Face ID, and you get back the consensus answer and how strongly they agreed. Use it for judgment calls where being wrong is costly: evals of AI output, tone and intent, whether two things match, moderation edge cases. Send refunds, payments and other money calls with route team. Write the question so a stranger can answer it from the context alone, and never include secrets or more personal data than the decision needs. Live keys use prepaid credit at $0.20 per answer (clips cost more by length), or $0.05 per internal vote with route team; test keys are free and simulated.
- **ask_judgment** — Ask a panel of ID verified people for their honest view when there's no right answer to check: which of two replies, designs, headlines or model answers is better, whether a customer would accept a reply or decision, how well something meets a brief (1 to 5), or whether a design, video or generated image has a flaw. You get the spread of views, a contested flag when people genuinely split, and a sample of one line reasons. format signoff (coming soon; test keys now) asks one top rated verified person to approve or reject an action after a fresh Face ID or Touch ID check, with a written reason: use it before an action someone must be accountable for. Live keys use prepaid credit at $0.20 per answer (a panel of 5 is $1.00) or $1.00 per sign off; test keys are free and simulated.
- **audit_my_agent** — Sends the developer's free Plures audit: up to 50 decisions an AI agent already made, each checked by ID verified people. Once per account, free with a test or live key. Each row is input (what the agent saw, including the policy it should follow) and decision (what it did). Gather them from the developer's logs, database or test fixtures; leave out names and other private details (emails, phones and card numbers are removed anyway). kind check (default) asks whether each call was right; kind judgment asks whether a customer would be happy with it. A person at Plures reviews it before anyone grades it, and results take a few days. Use get_audit_report to read them.
- **get_audit_report** — Reads the developer's Plures audit: where it is, and once people are answering, how many decisions they called right, wrong, unclear or split, with the ones they disagreed with listed first. For judgment audits: right means people would be happy with it, wrong means not happy, unclear means it needs changes.
- **find_precedents** — Searches the developer's judgment graph: past decisions people made on cases like this one, closest first, with the answer and how strongly they agreed. Returns a suggestion only when the case is settled: several close cases with the same options, all decided the same way with strong agreement. Free. To act on a settled case and have it recorded and billed as a decision, call ask_humans with use_precedent instead.
- **get_human_answer** — Check on a task sent with ask_humans and return the verified people's answer once it's ready.
- **cancel_human_task** — Cancel a task that is still waiting for answers, for example when the decision no longer matters. Answers already given are still billed.
- **get_mapping_guide** — Returns the steps for mapping the developer's AI agent for Plures: how to find its steps and actions, where to suggest a person checks, and how to draft each check. Call this first when asked to map an app for Plures.
- **submit_agent_map** — Uploads the map you made with get_mapping_guide: nodes, edges and checkpoints with drafted checks. Never include source code, secrets or customer data. If Plures names a problem, fix it and send again.
- **get_agent_map** — Returns an app's agent map from Plures: its checkpoints, which are added, and live numbers for each (checks in 30 days, agreement, how many went to the team).

## Claude Desktop Quick Installation
Install path detected from listing signals. Uses `npx` (confidence: high):

```json
"mcpServers": {
  "plures-mcp": {
    "command": "npx",
    "args": ["-y","plures-mcp"],
    "env": {
      "PLURES_API_KEY": ""
    }
  }
}
```

**Requires environment variables:** `PLURES_API_KEY` — the values above are empty placeholders; fill in real credentials before running (see the repository for what each one is for).

## Documentation

## What plures-mcp does

plures-mcp gives MCP-compatible agents access to Plures human verification workflows. An agent can submit a question, possible answers, and supporting context for review by multiple people whose identities are verified through government ID and a selfie check. The service returns the selected answer along with the strength of agreement and the individual votes.

The server exposes two tools. `ask_humans` is intended for decisions with defined options, such as checking whether an answer follows policy or interpreting a customer's tone. `ask_judgment` supports preferences, ratings, and sign-offs. For decisions involving money or accounts, requests can be routed to the user's own team with `route: "team"`.

## How it works

The agent sends Plures the question, options, and material to evaluate. Three verifiers are used by default, and a request can involve up to nine. They answer independently, with each response confirmed by Face ID, a fingerprint, or a passkey. When responses are split, Plures can ask additional people until a majority is reached or the configured limit is exhausted.

Results include an answer, an `agreement` value from 0 to 1, and the votes. A client can wait for the result or use a webhook-based flow in the broader Plures API. If a team is configured and the reviewers do not reach agreement, Plures can email the team a decision link and return the team decision as the final answer.

## Setup and configuration

Run the MCP server with `npx` and provide the account key through `PLURES_API_KEY`:

```json
{
  "mcpServers": {
    "plures": {
      "command": "npx",
      "args": ["-y", "plures-mcp"],
      "env": { "PLURES_API_KEY": "pk_test_..." }
    }
  }
}
```

The same configuration pattern can be used with Claude Desktop, Claude Code, Cursor, and VS Code. A free test key is available from Plures; test keys return immediate responses without contacting real reviewers, which supports development of the integration flow.

API keys belong to a Plures account rather than an individual. Account members can be assigned owner, developer, or viewer access in the Plures dashboard, while the key remains usable when a member leaves.

## Tools and capabilities

- Submit defined-option decisions to `ask_humans` for independent human answers.
- Request preferences, ratings, and sign-offs with `ask_judgment`.
- Include context or material for reviewers to assess.
- Receive consensus, agreement strength, and votes.
- Escalate unresolved or selected decisions to a configured team.
- Use test keys for immediate, non-production responses.

## Limitations and notes

Human review is not necessarily unanimous: requests can return a split result or expire, so agents should define how to handle insufficient agreement. Reviewers are selected and verified by Plures, and they do not see agent metadata. The normal reviewer count is three, with a maximum of nine. Team routing depends on having a team configured in the Plures dashboard.

Production integrations can use webhooks rather than waiting for a response, but webhook deliveries must be checked with Plures' `verifyWebhook` function in JavaScript or `verify_webhook` in Python. The MCP package itself is launched through `npx` and requires a Plures API key.

_Full upstream README: https://allmcps.com/mcp/plures-mcp/readme_