Name: Pugi
Author: pugi

Pugi MCP

AI product team: builds and publishes websites from a brief, logos, security checks.

Tool-list grade A Remote · Streamable HTTP API key 12 tools v1.0.0 MIT

No reviews yet by pugiGitHub The publisher’s own endpoint — this marketplace is not in the path

93/100

tool-list score

tools declared

remote

at its publisher

—

calls last 7d

What it does

AI product team: builds and publishes websites from a brief, logos, security checks.

Quickstart

# 1 — add it to Claude Code
claude mcp add --scope user --transport http pugi-mcp https://api.pugi.io/api/agent/v1/mcp --header 'Authorization: Bearer ${PUGI_AGENT_KEY}'

# 2 — ask your agent for something these tools do
#   get_balance, buy_credits_link, scan_url and 9 more

It needs PUGI_AGENT_KEY (Your Pugi agent key. Create one with POST https://api.pugi.io/api/agent/v1/keys (see https://pugi.io/agents.md)): this sends it in the Authorization header from your environment, so export it first — its endpoint answers 401 without it.

Pugi costs nothing on mcprush: there is no plan to choose here, no cap to set and nothing mcprush can bill you for.

Publishernot claimed

pugi1 server

1 category

All 1 server

Collected from a public index. Nobody has claimed this account, so nothing here was written by its author — claim it if it is yours.

Installation

Where are you running it?

Every route below installs the same thing and ends at the same approval screen. Nothing here runs on your machine and nothing runs on ours: this server answers at its publisher’s own address, and what you install is the connection straight to it.

Claude Codedeclared by this listing

This is a public server: it answers at its publisher’s own address and this marketplace is not in the path. Claude Code registers it in one command; --scope user makes it available in every project. The transport below is http, as the publisher's own manifest declares it. Claude Code reads PUGI_AGENT_KEY (Your Pugi agent key. Create one with POST https://api.pugi.io/api/agent/v1/keys (see https://pugi.io/agents.md)) from your environment each time it connects and sends it in the Authorization header — its endpoint answers 401 without it — so the key stays out of its config: export it in your shell profile (macOS, Linux), or run setx PUGI_AGENT_KEY … and open a new terminal (Windows).

claude mcp add --scope user --transport http pugi-mcp https://api.pugi.io/api/agent/v1/mcp --header 'Authorization: Bearer ${PUGI_AGENT_KEY}'

Reconnect, or start a new session, and the tools appear in the model’s tool list.

Free on mcprush, and it answers at its publisher’s own address: this marketplace is not in the path, so there is no key of ours to hold and no card to add — but its publisher asks for a key of their own on every call.

What it asks of you

It reaches a system of yours, so it needs your own credential rather than the publisher’s: Authorization header — Your Pugi agent key. Create one with POST https://api.pugi.io/api/agent/v1/keys (see https://pugi.io/agents.md).. You set it in your own client’s config, on your machine — this marketplace never holds it.

What it writes

One config entry pointing straight at the publisher’s own address — this marketplace is not in the path, and nothing from this listing executes on your machine.

What you approve

12 tools, with what each one reads, writes and reaches shown before you agree — the same list on every route above. Read the tool surface.

Tool surface

What the model actually sees. Descriptions are diffed on every release — see version history.

Only tools that write or reach the network

Shows the credit balance of this account in USD, what this key spent today, its daily spend cap, and the price of each tool. Paid tools spend this balance; when it runs out, ask a person to add credit with buy_credits_link, or pay with add_credit (a business purchase).

No parameter schema on file.

Returns a card checkout link for this account. A person opens it and pays there. This tool never charges anything by itself. Tax, where it applies, is added at checkout.

Parameter	Type	Required	Notes
amount_usd	number	no	How much credit to add. Rounded up to a pack of 5, 10, 25, 50 or 100 USD. Default 5.

A free, passive, read-only check of a public website: HTTPS and the redirect to it, five security headers, cookie flags, and the page's own scripts read for leaked keys. Returns a grade from A+ to F, counts by severity and the most serious finding. 10 checks a day per key. It is not a pentest and does not prove a site is secure.

Parameter	Type	Required	Notes
url	string	yes	The public address to check, for example https://example.com

Draws the same four whole logos as logo.pugi.io using its brief, concept planner and image inspection. Pass the user's description and style words verbatim, including unusual words; do not replace them with presets. Each logo includes a brand_kit_url opening that exact logo. Paid plans: included, up to 50 calls/day. Free accounts: three calls/day, then $0.25 per successful call. charged_usd and pr

Parameter	Type	Required	Notes
business_name	string	no	The business name, as written by the user.
industry	string	no	Optional. What the business does, for example "coffee shop" or "accounting firm".
description	string	no	The user’s description verbatim, including location, imagery and mood.
name	string	no	—
niche	string	no	—
tagline	string	no	—
styles	array	no	The user’s style words verbatim.
palette	string	no	—
refines	array	no	—
refine_notes	string	no	—
colours		no	Optional. Hex colours: {"primary": "#1e3a8a", "secondary": "#f59e0b", "text": "#111827"}, or a list of one to three of them in that order.
style	string	no	Style words verbatim; styles also accepts a list.
idempotency_key	string	yes	A unique string for this call, reused only to retry it. The same key with the same arguments is never charged twice.

Lists the apps this account has published, newest first, with their status and live address. Pass project_id to see one project.

Parameter	Type	Required	Notes
project_id	string	no	Optional project id.

Free. Searches the public Pugi catalogue of published sites and apps people built with Pugi. Filter by words in the title or description, tags (a card must carry all of them) and one category. Each result has a title, a short description, tags, a preview image, its public page on pugi.io and a start_from_this_url: the Pugi console with a brief for a new project like it (a person signs in there; it

Parameter	Type	Required	Notes
query	string	no	Optional words to find in the title or description, for example "booking".
tags	array	no	Optional tags; a card must carry every one, for example ["saas"].
category	string	no	Optional category.
limit	integer	no	How many cards to return. Default 10.

Checks a public GitHub repository branch for leaked secrets, vulnerable or malicious dependencies and risky code, and returns each finding with its severity, rule, file and line (or package) and a short fix hint, plus a summary. No code is run. $1 from prepaid credit, charged only when the check finishes; a failed check costs nothing. Pass an idempotency_key: a check can take a few minutes, and wh

Parameter	Type	Required	Notes
repo_url	string	yes	A public GitHub repository, for example https://github.com/owner/repo
ref	string	no	Optional branch name. Default: the repository's default branch.
idempotency_key	string	yes	A unique string naming this check. Reuse it to collect the result or to retry; the same key is never charged twice.

App builds require application security verification and must be started in the Pugi console. This tool returns app_builds_in_console without charging.

Parameter	Type	Required	Notes
logo_id	string	no	Selected logo_id returned by make_logo.
brandHandoffId	string	no	BrandHandoff from logo.pugi.io.
pick	string	no	Console gallery starter slug.
theme	string	no	Console theme slug.
locale	string	no	Site language.
brief	string	yes	—
max_usd	number	no	—
idempotency_key	string	yes	—

Builds a website from a brief in a Pugi workspace and returns a job_id. Billed by usage, never above max_usd (default 5, from 1 to 25 USD). max_usd is reserved from your credit before the build starts and the unused part is released when it ends; a build that produces nothing costs nothing. The answer includes estimate_usd_range from recent builds. A build takes several minutes: the call waits up

Parameter	Type	Required	Notes
logo_id	string	no	Selected logo_id returned by make_logo.
brandHandoffId	string	no	BrandHandoff from logo.pugi.io.
pick	string	no	Console gallery starter slug.
theme	string	no	Console theme slug.
locale	string	no	Site language.
brief	string	yes	What the site is for, who it is for, the pages and sections it needs, and any texts, colours or style it should use.
max_usd	number	no	The most this build may cost, in USD. Default 5. You pay the actual cost, usually less.
idempotency_key	string	yes	A unique string naming this job. The same key returns the same job and is never charged twice.

Makes one change in the workspace of a site this account built with build_site and returns a job_id. App edits must be made in the Pugi console. Billed by usage, never above max_usd (default 2, from 0.5 to 25 USD); a change that touches nothing costs nothing. Follow it with job_status like a build.

Parameter	Type	Required	Notes
logo_id	string	no	Selected logo_id returned by make_logo.
brandHandoffId	string	no	BrandHandoff from logo.pugi.io.
pick	string	no	Console gallery starter slug.
theme	string	no	Console theme slug.
locale	string	no	Site language.
job_id	string	no	The job_id of the build (or of an earlier edit) of the site to change.
project_id	string	no	Or the project_id of that site.
instruction	string	yes	The change to make, in plain words.
max_usd	number	no	The most this change may cost, in USD. Default 2.
idempotency_key	string	yes	A unique string naming this job. The same key returns the same job and is never charged twice.

Free. The state of a build_site, build_app or edit job: running, finished, stopped (at its spending limit, what was built is kept) or failed; a short progress line; charged_usd so far (final once the job ends); and, when it is built, a view-only preview_url that only shows the site and expires after 15 minutes (call again for a fresh one).

Parameter	Type	Required	Notes
job_id	string	yes	The job_id build_site, build_app or edit returned.

Free. Puts a site this account built live on its pugi.app address, after the same security check of the exact code every Pugi publish runs. The account's hosting rules apply (the number of live apps it may have). Answers the live url, or status "building" while the address comes up (then use publish_status).

Parameter	Type	Required	Notes
job_id	string	no	The job_id of the build (or the latest edit) to publish.
project_id	string	no	Or the project_id of that site.
idempotency_key	string	no	Optional. Reuse it to retry the same publish.

Free

$0 forever

Every tool — any call limit is its publisher’s, not ours
No card, and mcprush charges nothing for it
Runs at its publisher’s own address — this marketplace is not in the path
Nothing to cap, because mcprush bills nothing

Nothing here is billable: mcprush charges nothing to install Pugi or to call it, at any volume.

ATool-list grade · score 93 / 100

Release 1.0.0 · scanned 30 Oct 2026

Graded from the tools this server lists, not from its source code.

Two independent axes, because powerful and malicious are different questions. The grade is threat only. The capability level is blast radius, and it is never a penalty on the grade — it is priced as one subtract-only term in the score, where you can see it.

Threat score100

Client score93

Findings0 threat · 3 capability

Capabilitymoderate

Coveragemanifest

Verificationnot in this report

Surface digest52378eb0f58969cf7e85821f81bb9ff4c0e284d2bfb72fcadea618273dc1a36e

Score vectorthreat 100 → client 93

Term	Level	What it prices	Points
capability-exposure	moderate	capability blast radius (moderate) — client exposure if the model is manipulated	−3
coverage-honesty	manifest	inspection depth (manifest) — how much of the target the scan could see	−4

score = clamp(round(threatScore − E_cap − E_ver − E_cov), 0, 100). No model, no opinion, no network call at scoring time: the same methodology version against the same bytes always produces this number, and every line above is a field of the report rather than a summary of it.

What the scan could actually read

A grade is only as meaningful as its coverage, so the scanner publishes its own depth before it publishes its result.

The scan read each tool’s name, description, input schema and annotations exactly as the server listed them the last time it answered us; no code was reviewed. We ask the server again about once a month, and a list that changed is graded again.

Tool surfaceread

Implementation sourcenot available

Package metadatanot available

Live transportnot available

The implementation source was not read — this reflects the declared tool metadata only.

Capability — what it could do if the model were manipulated

Tags derived from each tool’s schema and the implementation, not from what the tool calls itself. moderate is the level these add up to.

Tool	Capability tags	Why the tag was assigned
get_balance	no tags
buy_credits_link	no tags
scan_url	untrusted-input	untrusted-input: declares openWorldHint: true (interacts with entities outside the server)
make_logo	external-sink	external-sink: declares readOnlyHint: false, openWorldHint: true
publish_status	no tags
catalog_search	no tags
scan_repo	external-sink	external-sink: declares readOnlyHint: false, openWorldHint: true
build_app	external-sink	external-sink: declares readOnlyHint: false, openWorldHint: true
build_site	state-write	state-write: declares readOnlyHint: false
edit	state-write	state-write: declares readOnlyHint: false
job_status	no tags
publish	external-sink	external-sink: declares readOnlyHint: false, openWorldHint: true

Toxic-flow graph

The lethal trifecta, checked as a graph rather than as a checklist: untrusted input, a sensitive source and an external sink have to meet before there is a path worth worrying about.

flow-1

Untrusted input can drive an external action even though no sensitive source is exposed.

Supply chain and provenance

Integrityfirst-seen

Surface digest52378eb0f58969cf

Scanned2026-10-30

This is the first scan of this surface here, so there is nothing yet to compare it against.

Every result on this tab comes from one deterministic pass over the published package — offline, rule by rule, and auditable line by line above. Same methodology version, same bytes, same score.