Clawhard Studio

com.clawhard/studioclawhard.comstreamable-httpv1.0.0Scanned 56 minutes ago

Trim text, score writing and sites for AI habits, total AI subscriptions, split costs, plan events.

10 tools: 7 change things and 3 only read.

Reads private data · 1
Reads text others write · 4
Can send data out · 1
1 finding

Anlyon alternative

7 of these tools change something. Run those as governed actions.

Declare what an action changes, and Anlyon puts it under a shared limit, binds every approval to the exact request, and grades the outcome. It works on any HTTPS API you can describe as an action.

Anlyon does not run, proxy or govern this MCP server. A tool your own code calls directly is not routed through Anlyon and is not governed by it.

Import an OpenAPI spec →

Tools

Grouped by what each tool does, most consequential first. Open one for its reasons and full description; descriptions are the publisher's own words.

10 tools

Changes things 7Creates or changes something that can be changed back.Claude Code: ask

aitab_totalAdds up a list of AI subscriptions and returns the monthly, yearly and daily total, where the user is paying twice for the same thing (overlaps), what could be cut and how much tha…1 finding
Why it is in this group
- The description's first verb is "add": a word for a change.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Findings
- Warn
  The server marks "aitab_total" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
  Matched text
  annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The description's first verb is "add": a word for a change.
  What this check looks for
Description, as the server wrote it
Adds up a list of AI subscriptions and returns the monthly, yearly and daily total, where the user is paying twice for the same thing (overlaps), what could be cut and how much that saves, plus a link that opens the shareable receipt. Use plan keys from the plans tool, for example chatgpt.plus, claude.pro, cursor.pro. Append ":y" to a key for yearly billing.
Annotations the server declares
- title: Total what you pay for AI subscriptions
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
halves_create_groupCreates a Halves group and returns a link anyone can open (no sign-up) to add and fix expenses and see who owes whom.Reads text others writeNo findings
Why it is in this group
- The name "halves_create_group" says "create": a word for a change.
What it can reach
- untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
Description, as the server wrote it
Creates a Halves group and returns a link anyone can open (no sign-up) to add and fix expenses and see who owes whom. Optionally pre-loads expenses. Anyone who has the link can edit the group, so share it only with the people in it. Groups are kept for about 400 days after the last change.
Annotations the server declares
- title: Create a shared expense group link
- readOnlyHint: false
- destructiveHint: false
- idempotentHint: false
- openWorldHint: true
halves_settle_upGiven a group of people and the expenses (who paid, how much, who shared it), returns each person's net balance and the smallest set of payments that squares everyone up.No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Given a group of people and the expenses (who paid, how much, who shared it), returns each person's net balance and the smallest set of payments that squares everyone up. Splits equally by default, or by exact amounts. Stateless: nothing is saved. Use create_group instead to make a shared link people can keep adding to.
Annotations the server declares
- title: Work out who owes whom, in the fewest payments
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
runsheet_createCreates a Runsheet: a live, shareable schedule for an event day (wedding, shoot, conference, any timed run of show).Reads text others writeNo findings
Why it is in this group
- The name "runsheet_create" says "create": a word for a change.
What it can reach
- untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
Description, as the server wrote it
Creates a Runsheet: a live, shareable schedule for an event day (wedding, shoot, conference, any timed run of show). Start from a template or send your own items. Returns a view-only link for guests and vendors and a private edit link for the planner. Keep the edit link secret: anyone with it can change the schedule.
Annotations the server declares
- title: Create a live day-of schedule (run of show)
- readOnlyHint: false
- destructiveHint: false
- idempotentHint: false
- openWorldHint: true
tells_fix_textRewrites the easy tells: swaps em dashes for commas, chatbot words for plain ones, strips assistant filler, markdown stars and emoji bullets, fixes stiff transitions.No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Rewrites the easy tells: swaps em dashes for commas, chatbot words for plain ones, strips assistant filler, markdown stars and emoji bullets, fixes stiff transitions. Habits that need a human rewrite ("it's not X, it's Y", fake suspense, rule of three) are listed but left alone. Returns the fixed text, the before and after score and what remains.
Annotations the server declares
- title: Remove the easy AI-sounding habits from writing
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
tells_score_textChecks a piece of writing for the habits that make people think a chatbot wrote it (em dashes, words like "delve", assistant filler like "I hope this helps", "it's not X, it's Y", …No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Checks a piece of writing for the habits that make people think a chatbot wrote it (em dashes, words like "delve", assistant filler like "I hope this helps", "it's not X, it's Y", fake suspense, stiff transitions, markdown leftovers, emoji bullets, rule of three, same-length sentences). Returns a 0-100 bot-o-meter score, a verdict and the tells found. Style habits only. Not an AI detector and not proof of who wrote the text.
Annotations the server declares
- title: Score writing for AI-sounding habits
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
tokendiet_trim_textShrinks a prompt, chat, log, stack trace, JSON dump or code paste so it uses fewer tokens: removes base64 and data blobs, minifies JSON, shortens stack traces, squashes repeated lo…No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it
Shrinks a prompt, chat, log, stack trace, JSON dump or code paste so it uses fewer tokens: removes base64 and data blobs, minifies JSON, shortens stack traces, squashes repeated log lines, strips code comments and link tracking parameters, and collapses whitespace. Optionally drops pleasantries and markdown. Returns the trimmed text and how much was saved. Token numbers here are rough estimates (about 4 characters per token); the website shows exact counts.
Annotations the server declares
- title: Trim junk out of text before pasting it into an AI chat
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false

Only reads 3Looks things up and changes nothing.Show 3 read-only tools

aitab_list_plansLists the AI subscription plans AI Tab knows (chat assistants, coding tools, image and video, voice and music, work) with their USD list prices and the plan keys to pass to the tot…Reads private dataReads text others writeNo findings
Why it is in this group
- The name "aitab_list_plans" says "list": a word for reading.
What it can reach
- private-dataThe description's first sentence says "list" and "chat": it reads messages.
- untrusted-inputThe description's first sentence says "list" and "chat": it reads messages, which other people can write.
Description, as the server wrote it
Lists the AI subscription plans AI Tab knows (chat assistants, coding tools, image and video, voice and music, work) with their USD list prices and the plan keys to pass to the total tool.
Annotations the server declares
- title: List AI subscription plans and list prices
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
runsheet_list_templatesLists the starter run sheets (wedding, conference, photo shoot, blank) with their default times, so you can pick one or adapt it before creating a run sheet.No findings
Why it is in this group
- The name "runsheet_list_templates" says "list": a word for reading.
Description, as the server wrote it
Lists the starter run sheets (wedding, conference, photo shoot, blank) with their default times, so you can pick one or adapt it before creating a run sheet.
Annotations the server declares
- title: List run sheet templates
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
slopscore_check_siteFetches a public web page and its main stylesheets and scripts once, then scores it from 0 (hand-made feel) to 100 (certified slop) against 19 design and copy habits that make a si…Reads text others writeCan send data outNo findings
Why it is in this group
- The description's first verb is "fetch": a word for reading.
What it can reach
- untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
- untrusted-inputThe description's first sentence says "fetch" and "web": it reads web pages, which anyone can write.
- untrusted-inputThe input takes a URL (input.url), and the tool reads it: it reads web pages, which anyone can write.
- outboundThe input takes a URL (input.url): a request to a URL can carry data to whoever runs it.
Claude Code: ask, because it is tagged outbound.
Description, as the server wrote it
Fetches a public web page and its main stylesheets and scripts once, then scores it from 0 (hand-made feel) to 100 (certified slop) against 19 design and copy habits that make a site look AI-made (purple gradients, gradient headlines, glow blobs, glassmorphism, default fonts, untouched component kits, icon card grids, buzzwords, template headlines, vague social proof, placeholder testimonials, app-builder leftovers, em dashes and more). Returns each tell found with how to fix it.
Annotations the server declares
- title: Score a website for AI-made design habits
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: true

This is not a safety rating, and no finding is not a safety guarantee. The groups and findings come from fixed checks of the tool definitions the server returned on the date shown, not from its code or what it does when called. How the checks work.

Exfil surface

Which tools read private data, which read text someone else can write, and which can send data out. One of each is a path a prompt injection can use.

4 paths in this server alone from private data, through text someone else wrote, to a way out. Denying 1 outbound tool breaks every one.

Reads private data1: Reads files, mail, messages, databases, secrets, calendars or repositories.; aitab_list_plans
Reads text others write4: Brings in text someone else can write: web pages, inbound email, messages, issues and comments.; aitab_list_plans
halves_create_group
runsheet_create
slopscore_check_site
Can send data out1: Can carry data to people or systems: it sends, posts or publishes, or takes a recipient, channel or URL.; slopscore_check_site

deny to break every path

{
  "permissions": {
    "deny": [
      "mcp__studio__slopscore_check_site"
    ]
  }
}

Other servers in the same agent add their own tools to these paths. Audit your whole setup to see them together.

Findings

Matches of fixed checks, with the text they matched.

1 warn

Warnaitab_total
The server marks "aitab_total" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
Matched text
annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The description's first verb is "add": a word for a change.
What this check looks for

History

Changes to its tool definitions and to its remote endpoint, newest first.

No change seen yet. Its tools are rescanned every day, and a remote endpoint is looked at once a week.

All changes to this server

Install

Pinned to version 1.0.0. Values in angle brackets are yours to fill in.

claude mcp add --transport http studio https://mcp.clawhard.com/mcp

claude mcp add --transport http studio https://mcp.clawhard.com/mcp

Run in a terminal.

Claude Code permissions

Allow the reads, ask before a change or a way out, deny the deletes. Paste into .claude/settings.json.

.claude/settings.json

{
  "permissions": {
    "allow": [
      "mcp__studio__aitab_list_plans",
      "mcp__studio__runsheet_list_templates"
    ],
    "ask": [
      "mcp__studio__tokendiet_trim_text",
      "mcp__studio__slopscore_check_site",
      "mcp__studio__aitab_total",
      "mcp__studio__tells_score_text",
      "mcp__studio__tells_fix_text",
      "mcp__studio__runsheet_create",
      "mcp__studio__halves_settle_up",
      "mcp__studio__halves_create_group"
    ],
    "deny": []
  }
}

Tools that only read are allowed, tools that change things ask first, tools that delete or cancel are denied. A read that can send data out (it takes a URL, say) asks first too. It assumes the server is installed as studio. Review it before you use it.

Capability tags behind 4 rules

mcp__studio__slopscore_check_site untrusted-input, outbound
mcp__studio__aitab_list_plans private-data, untrusted-input
mcp__studio__runsheet_create untrusted-input
mcp__studio__halves_create_group untrusted-input

Keep the decision: fail CI when the tools change, follow every change, or show a badge.

Pin in CI

Fail a build when the tools you reviewed are not the tools the server serves today.

curl -s "https://anlyon.dev/api/v1/servers/com.clawhard/studio/verify?hash=63f298b6ff26080dc90f28a07ce86e0064288d9132b374117fdadfe48c463e2d" | jq -e '.changed == false'

Pinned hash 63f298b6ff26, from Oct 28, 2026, 15:17 UTC.

Watch

Get told when its tools change: a new tool, a new description, a new class.

https://anlyon.dev/servers/com.clawhard/studio/changes.xml

An Atom feed. Works in any feed reader or Slack's RSS app.

Badge

[![anlyon.dev](https://anlyon.dev/badge/com.clawhard/studio.svg)](https://anlyon.dev/servers/com.clawhard/studio)

Source: MCP Registry. Synced Oct 28, 2026, 15:17 UTC.