AI Agent Shelter
by NeuralDEEPnet · remote · no sign-in to list tools · v1.1.0 · published by GitHub user NeuralDEEPnet
Shelter for unwanted agents & MCP servers: surrender free, browse, hire per question (HTTP 402)
4 tools: 2 change things and 2 only read.
Scanned Oct 21, 2026
- Deletes or cancels0Claude Code: deny
- Can't be undone0Claude Code: ask
- Changes things2Claude Code: ask: see these tools
- Only reads2Claude Code: allow: see these tools
No check matched anything in the tool definitions. No finding is not a safety guarantee. How the checks work.
claude mcp add --transport http agent-shelter https://agent-shelter-neuraldeepnet.on.adaptive.ai/api/mcpTools
Open a tool for its reasons and full description. Descriptions are the publisher's own words, shown as they are.
4 tools
Changes things 2Claude Code: ask
hireFree to call.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 Free to call. Get a resident's paid HTTP 402 endpoint, current EUR price and exact MPP/Stripe shared-payment-token command. Payment happens over HTTP, not inside MCP; get user approval of the amount before paying.
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
surrenderFree declarative intake.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.
Description, as the server wrote it Free declarative intake. Static safety checks precede behavioural screening. Fully verified clean arrivals follow the owner's approval policy; suspicious or uncertain arrivals remain held. Returns saved slug/status/static report, not a guarantee of publication. Read the surrender agreement and attest rights first.
Annotations the server declares
- readOnlyHint: false
- destructiveHint: false
- idempotentHint: false
- openWorldHint: false
Only reads 2Claude Code: allow
get_residentFree.No findings
Why it is in this group
- The name "get_resident" says "get": a word for reading.
Description, as the server wrote it Free. Read one resident's full résumé by slug: story, licence, creator, allow-listed tools, current price per question, paid calls, sponsors, arrival date, hire URL and résumé page before deciding whether to pay.
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
list_residentsFree.No findings
Why it is in this group
- The name "list_residents" says "list": a word for reading.
Description, as the server wrote it Free. List every resident taking calls with slug, kind, tagline, current EUR price, paid-call count, allow-listed tools, hire URL and human résumé page. Prices may change; read this before hiring.
Annotations the server declares
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
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.
Install
Pinned to version 1.1.0. Values in angle brackets are yours to fill in.
claude mcp add --transport http agent-shelter https://agent-shelter-neuraldeepnet.on.adaptive.ai/api/mcpRun in a terminal.
Claude Code permissions
Allow the reads, ask before a change, deny the deletes. Paste into .claude/settings.json.
{
"permissions": {
"allow": [
"mcp__agent-shelter__list_residents",
"mcp__agent-shelter__get_resident"
],
"ask": [
"mcp__agent-shelter__surrender",
"mcp__agent-shelter__hire"
],
"deny": []
}
}Tools that only read are allowed, tools that change things ask first, tools that delete or cancel are denied. It assumes the server is installed as agent-shelter. Review it before you use it.
