cite-gate MCP server
online · 100% uptimekr.co.mlpc/cite-gate·v0.1.0Checks that every quote in an AI-written script appears word for word in its source.
0 stars
Reviews
Write oneNobody has reviewed cite-gate yet.
If you have run it, two minutes of your experience saves the next person an afternoon.
cite-gate tools (2)
write = sends, deletes, buys or postscheck_citationsFreeCheck that every factual item in a script is backed by a quote that appears word for word in a source text, that numbers and names in the prose appear in the quotes, and that the prose retells rather than copies. Hosted limit: 3000 characters of prose per call; for longer scripts run the local server.
check_numbersFreeList the numbers in a piece of prose that do not appear in any of the given quotes.
Public scan report
scanner v0.1.10 · 2026-10-13 · same rubric, same numbers if you re-run it
- –Code scanremote-only server, no package to scann/a
- Live reliabilityremote reachable in 481ms20/20
- Tool poisoning2 tool descriptions checked15/15
- Auth qualityopen endpoint, read-only tools10/15
- Maintenancelast push 4 days ago15/15
- Maintainer identityendpoint served from the registry-verified namespace domain8/10
How cite-gate compares
Against every graded MCP server, from our own scans and uptime checks. Updated after each scan run.
- Safety score
- 91/100
- higher than 85% of 48,255 servers
- Tools
- 2
- the typical server has 7
- Tools that change things
- 0%
- typical: 0%. Fewer write tools means less an agent can break
- Uptime
- 100%
- of our recent checks; the typical remote server answers 100%
- Response time
- 213 ms
- faster than 61% of 30,031 servers
What its tools let an agent do: 2 read
Install cite-gate in Claude Code, Cursor or VS Code
claude mcp add --transport http cite-gate https://mcp.mlpc.co.kr/mcp
What the publisher says
From the cite-gate repository's README, as published. We do not edit it. Read it on GitHub
cite-gate
A small checker for scripts written with AI help. It makes every factual item in a script carry a quote that appears word for word in a source file, and it flags prose that copies the source instead of retelling it.
It doesn't decide whether a claim is true. It decides whether you can point at the sentence your claim came from, and whether that sentence actually exists.
$ python -m cite_gate check examples/gettysburg/claims.json
5 items · 91 words · 2 failures · 1 warnings
FAIL [number-not-in-quote] number(s) ['1776', '87'] not in the quotes
FAIL [quote-not-in-source] quote not found in 'gettysburg': 'this nation shall endure forever and ever'
warn [copied-not-retold] 26 consecutive words shared with the quotes; retell, don't copyThe example fails on purpose: its last three items are mistakes that a language model makes all the time (a number it worked out itself, a quote it half remembered, a paragraph lifted from the source).
Why
Language models write fluent prose with confident details. The details that go wrong are the small ones: a year that's off by one, a name added "for context", a quote paraphrased into something the source never said. Reading the whole script against the sources by hand doesn't scale. This tool turns that reading into a gate you can run on every change.
What it checks
Exit code is 1 when there are failures. --strict also fails on warnings.
Input format
A claims file (JSON) and a folder of plain-text sources, one file per source id.
{
"config": {"allow": ["Lincoln"], "copy_run": 11},
"sections": [
{"title": "The occasion",
"items": [
{"id": "opening",
"text": "He began by looking back four score and seven years ...",
"claims": [{"src": "gettysburg",
"quote": "Four score and seven years ago our fathers brought forth on this continent, a new nation"}]}
]}
]
}sources/gettysburg.txt holds the source text. Keep your own copy of each source (with its URL and revision noted at the top), so a later run checks against the same text.
Config
Use
Command line (Python 3.10+, no dependencies):
pip install . # from a checkout; adds the `cite-gate` command
cite-gate check path/to/claims.json [--sources DIR] [--json] [--strict]
# or without installing
python -m cite_gate check path/to/claims.jsonGitHub Actions:
- uses: actions/checkout@v4
- uses: cite-gate/cite-gate@v0
with:
claims: script/claims.json
strict: "false"Use from your agent (MCP)
Shortened. The full README is on GitHub.
Nothing above is checked by us. What we check is on the safety report.
cite-gate: common questions
- Is cite-gate MCP server safe?
- Yes, by our scan: it is graded A (91/100). Read the cite-gate safety report
- How do I install cite-gate?
- It runs remotely at mcp.mlpc.co.kr. Add it to Claude Code, Cursor, VS Code or Claude Desktop with the snippets above, or call it through the mcp.market gateway without installing anything.
- Does cite-gate need an API key?
- Not as far as the registry entry and our scan can tell: no credentials are declared or required.
- Is cite-gate maintained?
- The last commit was 26 days ago (2026-10-09). The latest release is v0.1.0.
- Is cite-gate up?
- 100% of our last 28 checks got an answer. We check remote servers about four times a day.
- How does cite-gate compare with other MCP servers?
- safety score 91/100 (higher than 85% of 48,255 servers); tools 2 (the typical server has 7); tools that change things 0% (typical: 0%. Fewer write tools means less an agent can break). Full comparison.