free-agent-toolsSAFE
Free no-auth remote MCP server (streamable HTTP): 16 deterministic tools for business ideas, pricing, hotel OTA fees, SaaS savings, ads notice risk, App Store 4.2 precheck, Japan trips.
Overview
From the repository's own README, as read at the audited commit. Badges and raw HTML are left out.
A free, no-auth remote MCP server (streamable HTTP) with 19 small, deterministic tools your AI agent can call: startup pitch roasting, business idea scoring, price headroom, hotel OTA commission, open-source swaps for SaaS bills, Google Ads / Meta Ads notice risk, App Store 4.2 / 4.3 precheck, faceless YouTube reality check, AI infrastructure bottlenecks, viral attention patterns, Japan trip planning, Japan tax-free refund checking, and Japan visa / residence permit fees.
Endpoint: https://free-agent-tools.vercel.app/mcp Landing page with setup for every client: https://free-agent-tools.vercel.app
No API key, no account, nothing to install. Results are rules of thumb and educational estimates, not financial, legal, or investment advice, and every result carries a disclaimer for the user.
Add it
Claude Code
claude mcp add --transport http free-agent-tools https://free-agent-tools.vercel.app/mcp
Claude (desktop / web): Settings → Connectors → Add custom connector → paste the URL.
ChatGPT: Settings → Apps & Connectors → Advanced → Developer mode → Create → paste the URL, no authentication.
Cursor (~/.cursor/mcp.json)
{ "mcpServers": { "free-agent-tools": { "url": "https://free-agent-tools.vercel.app/mcp" } } }VS Code (.vscode/mcp.json)
{ "servers": { "free-agent-tools": { "type": "http", "url": "https://free-agent-tools.vercel.app/mcp" } } }Windsurf (~/.codeium/windsurf/mcp_config.json)
{ "mcpServers": { "free-agent-tools": { "serverUrl": "https://free-agent-tools.vercel.app/mcp" } } }Stdio-only clients
{ "mcpServers": { "free-agent-tools": { "command": "npx", "args": ["-y", "mcp-remote", "https://free-agent-tools.vercel.app/mcp"] } } }Tools
55ec06fc7559OBSERVED · 2026-10-11Connect
Built from this server's own package name, version and transport as found in its source — not copied from anyone's documentation, so it cannot drift against a page we do not control.
claude mcp add free-agent-tools -- npx -y [email protected]
{
"mcpServers": {
"free-agent-tools": {
"command": "npx",
"args": [
"-y",
"[email protected]"
]
}
}
}Exposed tools (14)
13 read · 1 write · 0 destructive.
| Tool | Risk | Description |
|---|---|---|
ai_bottleneck_quiz | read | Without answers, returns the five quiz questions with options. With answers (one slug per question from: ${BOTTLENECKS.join( |
ai_infrastructure_bottlenecks | read | Explains the physical constraints on the AI buildout beyond chips: compute, memory (HBM), optics, power, space (sites, backhaul), and servers (racks, cooling). Omit slug for all six. Includes example public companies often cited in discussion; these are not investment picks. |
app_store_wrapper_precheck | read | For Capacitor, WebView, PWA-shell, React Native, or AI-generated iOS apps: scores rejection risk under Guideline 4.2 (minimum functionality / web wrapper), 4.3 (spam / clone), and metadata, with reasons and how native each feature reads. Not legal advice; Apple decides. |
email_list_reactivation_value | read | Estimates buyers and revenue from one offer to past customers in low / mid / high scenarios, with an optional partner revenue share. |
faceless_youtube_reality_check | read | Seven multiple-choice answers about a faceless or AI YouTube channel plan. Returns HIGH / MED / LOW expectation and monetization-policy risk with signals and myths to drop. Pushes back on viral |
hotel_ota_commission_calculator | read | For hotels, ryokan, guesthouses, and B&Bs: yearly room revenue, commission paid to OTAs (Booking.com, Expedia, Agoda...), commission as a share of revenue (HIGH / MED / LOW), and net savings from moving a share of OTA bookings to direct. Any currency. |
japan_trip_options | read | Lists supported Japanese cities (with regions, peak months, and stay areas), months with season and crowd level, and traveler types for japan_trip_plan. |
list_open_source_saas_alternatives | read | Lists every covered paid tool (Mixpanel, Semrush, FreshBooks, Calendly, Chargebee, Typeform, Pipedrive/HubSpot, GoHighLevel, Intercom/Zendesk) with its open-source swap, GitHub repo, license, setup effort, and main catch. |
price_headroom_check | read | Is the business underpriced? Uses sales close rate (80%+ = way underpriced, ~30% = about right, under 25% = sales problem) to suggest a price range, and computes the profit multiple of a price rise after losing some customers, plus the break-even customer loss. |
saas_self_host_savings | read | Give monthly spend per paid tool (ids: ${SAAS_IDS.join( |
score_business_idea_moat | read | Score a business idea on Margin, Operations, Advantage, and TAM (1-10 each). 30+ = FUND IT, 20-29 = FIX IT, under 20 = FLEE IT. Returns the weakest factor, how to fix it, and red flags from pain / money / willingness-to-suffer checks. Educational rule of thumb, not financial advice. |
thirty_day_cash_check | read | Compares cash collected in a new customer |
viral_attention_patterns | write | Patterns of how attention spreads (TikTok sound reuse, X quote-post piles, stitch chains, group-chat forwards, and more), each with the signal to watch and the lesson. Optional platform filter. Educational; attention is not an investment signal. |
viral_attention_quiz | read | Without answers, returns six questions. With answers (one of chaser, inventor, reader per question), returns the user |
Trust audit
SAFEgrade B · trust 89/100 Nothing in the source contradicts what it says it does. Grade A is reserved for packages that have also passed the behavioural sandbox.
| Layer | What it checks | Result |
|---|---|---|
| L0 | Provenance & inventory | PASS |
| L1 | Static analysis of the code | PASS |
| L2 | Instruction surface (what it tells the agent) | PASS |
| L3 | Class-specific surface | WARN |
| L4 | Behavioural (sandbox) | SKIPPED |
What the source does
- Filesystem
- none-observed
- Network
- none-observed
- Shell
- none-observed
- Dependencies
- not all pinned
- Secrets in source
- none-found
Findings (2)
streamable-http
@modelcontextprotocol/server, mcp-handler, next, react, react-dom, zod, @modelcontextprotocol/client, @types/node
Gates applied: no_behavioural_pass.
55ec06fc7559full audit observations/trust-audit/mcp-server/sean0007__free-agent-tools.json · Report an issue / request a re-scanAudit history
Every audit this server has had. A grade with a past is a grade somebody is still checking.
| Date | Source | Verdict | Grade | Score | Change |
|---|---|---|---|---|---|
| 2026-10-11 | 55ec06fc7559 | SAFE | B | 89 | first audit |
Questions
What is the free-agent-tools MCP server?
Free no-auth remote MCP server (streamable HTTP): 16 deterministic tools for business ideas, pricing, hotel OTA fees, SaaS savings, ads notice risk, App Store 4.2 precheck, Japan trips.
What tools does free-agent-tools expose?
14 in total: 13 read-only, 1 that write, and 0 that can delete or overwrite. Every one is listed on this page with its risk.
Is free-agent-tools safe to connect to an agent?
The audit found nothing in the source that contradicts what it says it does, and graded it B (89/100). Grade A is held back for packages that have also passed a sandboxed behavioural run, which is why a clean server reads B.
What credentials does free-agent-tools need?
No credential environment variables were found in its source, so it appears to need none.
How does free-agent-tools run?
It speaks streamable-http, so it runs as a service you connect to over the network. It is published on npm as free-agent-tools at 1.3.0.
How current is this page?
The grade is for one exact copy of the source (55ec06fc7559), read on 2026-10-11. The repository is watched and re-audited when it changes.