sentinel-proof

io.github.92kareeemPublished by GitHub account 92kareeemv0.1.2Updated Nov 16, 2026

Checks an answer against its sources: numbers and citations verified, wrong ones fixed. No model.

The publisher's own server: Proof Holdings

VerifiedSTDIODesktop only

Overview

sentinel-proof's local MCP server; no sign-in required; STDIO; 1 tool.

No AI overview yet: these facts come from the registry entry and the server itself.

Installation

In SourceWeft

  1. Open sentinel-proof in the dashboard and add it to a workspace.
  2. Enable the server for the chats that should use its tools.

Desktop only via STDIO. STDIO servers start a local process, so they need the SourceWeft desktop host.

Other MCP clients

Follow the launch instructions in the repository.

README

sentinel-proof-mcp: stop your agent from showing hallucinated answers

[PyPI] [Python] [MCP registry] [Framework compatibility] [No API key] [License: Apache-2.0]

An MCP server that fact-checks an LLM answer against the documents it was written from, before the user sees it. Claude Desktop, Claude Code, Cursor, VS Code, Windsurf or any Model Context Protocol client gets one tool, verify_answer. The agent passes its retrieved passages and its draft; the tool checks that every number appears in the passage the sentence cites and that every citation points at a passage that supports it. Wrong citations are re-pointed, unsupported sentences are removed, and an answer that is mostly unsupported is refused with a reason. The agent then shows safe_text, not the draft.

Runs locally over stdio. No model call, no network, no API key, no GPU: the check is sentinel-proof, a deterministic grounding gate that takes microseconds, so it can run on every answer rather than a sample. Nothing you verify leaves your machine.

bash
claude mcp add --scope user sentinel-proof -- uvx sentinel-proof-mcp

Or try the same check in your browser first; nothing you paste is uploaded.

Use it when:

  • your agent answers from files, a vector store or search results and sometimes makes up a price, date, limit or percentage that is not in them;
  • it cites the wrong source, so a user who clicks [2] reads a passage that says nothing of the kind;
  • you want the answer grounded or refused, not a faithfulness score you then have to threshold, and not a second model call that is itself a model;
  • you need a check that is the same every time, for a regulated domain (legal, finance, healthcare, HR policy) or for a test suite;
  • you want it inside the agent loop, so the model fixes its own answer before replying, instead of a dashboard that tells you afterwards.

Install: works with every MCP client

MCP is an open protocol, and this server is a plain stdio process, so it is not tied to Claude or to any model. Any client or agent framework that can launch an MCP server can use it, with any LLM behind it. One prerequisite: uv (pipx works too). Every client below then launches the server on demand with uvx sentinel-proof-mcp; nothing else to install.

[Install in Cursor] [Install in VS Code]

One command, for clients with a CLI:

bash
claude mcp add --scope user sentinel-proof -- uvx sentinel-proof-mcp   # Claude Codecodex mcp add sentinel-proof -- uvx sentinel-proof-mcp                 # OpenAI Codex CLIgemini mcp add sentinel-proof uvx sentinel-proof-mcp                   # Gemini CLI

One JSON block, for clients with a config file. This is the standard mcpServers shape:

json
{  "mcpServers": {    "sentinel-proof": {      "command": "uvx",      "args": ["sentinel-proof-mcp"]    }  }}
ClientWhere the block goes
Claude Desktopclaude_desktop_config.json (Settings → Developer → Edit Config)
Cursor.cursor/mcp.json in the project, or ~/.cursor/mcp.json
Windsurf~/.codeium/windsurf/mcp_config.json
Cline, Roo Codethe MCP settings file the extension opens from its MCP panel
LM Studiomcp.json (Program → Install → Edit mcp.json)
ChatGPT desktop, Goose, Continue, Zed, JetBrains AI, Kiro, OpenCode, Warpthe same command and args; see each client's MCP settings for the file
VS Code (Copilot agent mode).vscode/mcp.json, with the top-level key servers instead of mcpServers

Over HTTP, for clients that connect to a URL rather than launching a process (and for Docker, a shared dev box, or an agent running elsewhere):

bash
uvx sentinel-proof-mcp --transport streamable-http --host 0.0.0.0 --port 8765# clients connect to http://<host>:8765/mcp

From Python, if you would rather manage the environment yourself: pip install sentinel-proof-mcp, then the command is sentinel-proof-mcp (or python -m sentinel_proof_mcp).

Windows note: if a desktop client cannot find uvx, put the full path in command (where uvx prints it) or set uv's install directory on PATH for GUI apps.

From an agent framework

The server works with any LLM provider, because the framework, not the server, talks to the model. These are run nightly in CI against the latest release of each framework:

OpenAI Agents SDK

python
from agents import Agent, Runnerfrom agents.mcp import MCPServerStdio
async with MCPServerStdio(name="sentinel-proof", params={"command": "uvx", "args": ["sentinel-proof-mcp"]}) as proof:    agent = Agent(name="docs-qa", instructions="Answer from the passages, cite [n], then verify_answer before replying.", mcp_servers=[proof])    result = await Runner.run(agent, "What is the refund policy?")

LangChain / LangGraph

python
from langchain_mcp_adapters.client import MultiServerMCPClient
client = MultiServerMCPClient({    "sentinel-proof": {"command": "uvx", "args": ["sentinel-proof-mcp"], "transport": "stdio"},})tools = await client.get_tools()   # [verify_answer], ready for create_agent / bind_tools

langchain-mcp-adapters pins the mcp package below 2.0 and this server needs 2.x. That is why the examples launch it with uvx, which gives it its own environment; the two versions talk over the wire without issue.

Pydantic AI

python
from fastmcp.client.transports import StdioTransportfrom pydantic_ai import Agentfrom pydantic_ai.mcp import MCPToolset
proof = MCPToolset(StdioTransport(command="uvx", args=["sentinel-proof-mcp"]))agent = Agent("<provider>:<model>", toolsets=[proof])   # any provider Pydantic AI supports

Google ADK, LlamaIndex, CrewAI, AutoGen, Semantic Kernel, Mastra, Vercel AI SDK and n8n all have an MCP client that takes the same command and args.

What the agent gets

One tool, verify_answer:

Argument
answerthe draft, with citations
passagesthe retrieved texts in the order the model saw them: strings, or {"text", "id", "source", "section", "page"}
questionoptional; figures the user typed are not counted as fabrications
citation_style"numeric" (default): [1], [2, 3] are 1-based positions in passages; "tag": [chunk:<id>]
max_stripped_ratiorefuse when more than this fraction of claims is stripped (default 0.5)

It returns structured JSON: ok, safe_text (what to show), summary (one line telling the model what happened and what to do), claims (each sentence with supported / repaired / stripped, what it cited, what supports it, and any figures no passage contains), repaired_citations, stripped_ratio, refusal.

json
{  "ok": true,  "summary": "Verified: 2 of 3 claims supported. 1 citation(s) re-pointed to the right passage. 1 unsupported claim(s) removed. Show the user safe_text.",  "safe_text": "Customers can get a refund within 30 days [1]. Requests above 500 USD need manager approval [2].",  "claims": [    {"text": "Customers can get a refund within 30 days", "status": "supported", "cited": ["p1"], "supported_by": ["p1"], "unsupported_numbers": []},    {"text": "Requests above 500 USD need manager approval", "status": "repaired", "cited": ["p1"], "supported_by": ["p2"], "unsupported_numbers": []},    {"text": "Refunds are processed in 2 business days", "status": "stripped", "cited": ["p2"], "supported_by": [], "unsupported_numbers": ["2"]}  ],  "repaired_citations": 1,  "stripped_ratio": 0.333,  "refusal": null}

One prompt, cite_for_verification, which tells the model how to draft so that every sentence can be checked: cite each sentence as [n], copy figures exactly, reply INSUFFICIENT_CONTEXT when the sources do not answer. Use it as the system instruction of any answer-from-documents step, and the tool will have something to verify.

One resource, sentinel-proof://about, with the versions and the rules.

A typical turn

  1. The agent retrieves passages (its own search tool, a vector store, files).
  2. It drafts an answer with [1]-style citations (the prompt above).
  3. It calls verify_answer(answer, passages, question).
  4. It shows safe_text. If ok is false, summary says what was wrong: fabricated figures, no citations, or too little support, and what to do next (answer only from the passages, retrieve more, or tell the user).

The server's instructions say exactly this, so a client that reads them will do it without further prompting. In Claude Desktop or Cursor you can also just say "verify that against the sources with sentinel-proof".

How it compares

Other ways to catch a hallucinated answer from inside an MCP agent, and where this one differs. None of these is wrong; they answer different questions.

Where it runsNeedsWhat you get backFixes the answer
sentinel-proof-mcpyour machinenothingper-sentence verdict, safe_text, refusal with reasonyes: re-points citations, drops unsupported sentences
Ask the model to check itselfthe modela second model callan opinion, different each runsometimes, unverifiably
Hosted NLI / "verify" serversa vendor's APIAPI key, your text uploadeda score or label per claimno
URL fact-checkersthe webnetworkwhether a page says Xno; and your passages are not URLs
RAG evaluation frameworksyour CIa judge modelfaithfulness scores over a datasetno; offline

The trade: this check is lexical and numeric, not semantic. It catches the invented figure and the citation to the wrong passage, which are the hallucinations users notice and screenshot. It does not catch a paraphrase that misstates its source using the source's own words; for that, put a judge model behind it. It is the deterministic floor, not the ceiling.

FAQ

Does it need an API key or send my documents anywhere? No. The server is a local process over stdio, the check is pure Python with no dependencies, and it makes no network calls. The passages and the answer stay in the process.

Which clients and models does it work with? Any MCP client, with any model. The server never talks to an LLM; the client does. It speaks stdio, the transport every desktop client and CLI uses to launch local servers, and streamable HTTP for the rest. The test suite drives it over stdio as a real subprocess, and the framework examples above were run against this release with the OpenAI Agents SDK, LangChain (on the older mcp 1.x) and Pydantic AI.

How is this different from asking Claude to double-check its answer? The model checking itself is another model call: it costs tokens, takes seconds, and gives a different answer each time. This tool is a rule: a number is in the cited passage or it is not. It returns the same verdict every run, in microseconds, and tells the model exactly which sentence failed and why.

My agent does not cite. Will it still work? Not yet: with nothing cited there is nothing to check, and the tool says so (ok: false, "No citations could be parsed"). Attach the cite_for_verification prompt to the drafting step, or add "cite each sentence as [n]" to your system prompt. Citation is what makes the answer checkable.

What counts as a passage? Whatever the model saw: chunks from a vector store, paragraphs from a file, search snippets, tool results. Pass them in the order the model saw them so that [1] means the first. Objects with id, source, section and page let the verdict say which document a sentence came from.

Does it work in languages other than English? Numbers are checked in any language. The word-overlap rule uses an English stop-word list, so for other languages it is stricter than intended; multilingual stop-words are on the roadmap of sentinel-proof.

Is there a version I can call from Python without MCP? sentinel-proof is the library; verify(answer, passages) is the same check. This package only puts it behind an MCP tool.

Where is it listed? In the official MCP registry as io.github.92kareeem/sentinel-proof, and on PyPI.

What it does not do

It checks evidence, not meaning: word overlap and numbers, no entailment model. A paraphrase that misstates its source and shares its vocabulary passes; an invented figure or a citation to the wrong passage does not. It is the deterministic floor under a judge model, not a replacement for one. Details and the design decision are in the sentinel-proof README.

Development

bash
pip install -e ../sentinel-proof -e ".[dev]"pytest -q            # in-process through the MCP client, plus one real stdio subprocessruff check . && mypy srcpython compat/check_openai_agents.py   # launch the installed server from a framework's MCP client

compat/ holds one check per agent framework. A nightly workflow runs them against the latest release of each framework and of this package, plus the official MCP Inspector, so the client list above stays true.

License

Apache-2.0.

Source: packages/sentinel-proof-mcp/README.md at commit 688aa57

Tools

1

verify_answer

Verify an answer against its sources

Check a drafted answer against the passages it cites, sentence by sentence. Every number in a sentence must appear in a passage it cites, and the sentence's words must overlap that passage. A sentence whose citation is wrong is re-pointed to the passage that does support it ("repaired"); a sentence no passage supports is removed ("stripped"); if more than max_stripped_ratio of the sentences are stripped the whole answer is refused and `safe_text` is a refusal instead. `passages` are the retrieved texts in the order they were shown to the model, as strings or as {"text", "id", "source", "section", "page"}. With citation_style "numeric" the answer cites them as [1], [2, 3] (1-based positions); with "tag" it cites [chunk:<id>]. `question` lets figures the user typed pass without being counted as fabrications. Deterministic: no model, no network. Returns the verdict with per-claim detail, a one-line `summary`, and `safe_text` to show.

Parametersanswerrequiredpassagesrequiredquestioncitation_stylemax_stripped_ratio

Version history

2
  1. v0.1.2LatestNov 16, 2026
  2. v0.1.1Nov 16, 2026

Score

90% of the applicable checks met

How complete this listing is, computed from the registry entry, the endpoint and public sources. Not a rating of the server and not a statement that it is safe or working today: reachability is a dated check, and no installs or reviews go into the score.

  • Reached in the last 30 days25 pts
  • Auth declared10 pts

    Declare the credential headers in the registry entry, or keep the endpoint answering the probe so its auth can be classified.

  • Tools listedRequired15 pts
  • DescriptionRequired15 pts
  • Logo10 pts
  • Documentation15 pts
  • License10 pts

Other Proof integrations