Cybersecurity
phishclean-mcp-guard
Audit and firewall for AI agent MCP server permissions
- 0community votes
- 32DR by Ahrefs no change since last check
- 2025founded
Ahrefs Domain Rating listed 20 Nov 2026, relisted 17 Nov 2026.
Free browser-based audit tool that shows what your AI agents can reach through MCP servers, identifies over-permissioned tools, and offers MCP Guard, a local
What it is

phishclean-mcp-guard: the homepage
The MCP Security Audit is a free browser-based tool that analyzes MCP (Model Context Protocol) configurations from Claude Desktop, Claude Code, Cursor, VS Code, and Windsurf. It inventories every MCP server and tool, identifies which ones have over-permissioned access, detects inline secrets and unpinned packages, and shows the "lethal trifecta" - when an agent can read private data, access content from outsiders, and send data out, creating a path for prompt injection attacks to steal information.
MCP Guard is an optional open-source firewall that runs locally on your machine as a proxy between your AI agent and each MCP server. It enforces read-only mode for databases and sensitive systems, blocks calls that try to exfiltrate secrets, pins tool definitions to detect "rug pulls" when updates change tool behavior, filters poisoned tool descriptions, flags injection attempts, and logs all activity. Installation is one command and policies can be shared with teams via a git repository.
Listed capabilities
- MCP server inventory and permissions analysis
- Over-permissioned tool detection
- Local config analysis in browser
- Read-only mode enforcement for databases and APIs
- Secret and API key leak blocking
- Tool definition pinning to detect rug pulls
- Prompt injection detection and flagging
- Activity logging and webhook alerts
- Team policy management via git repository
Pricing
Pricing model: Free
Follow phishclean-mcp-guard for an email about each milestone, update and new product they post, in a daily summary.
phishclean-mcp-guard metrics
| Metric | Value |
|---|---|
| Domain Rating | 32 no change since last check |
| Trust Flow | 0 no change since last check |
| Citation Flow | 19 no change since last check |
| Referring domains | 8 no change since last check |
| External backlinks | 17 no change since last check |
Ahrefs Domain Rating listed 20 Nov 2026, relisted 17 Nov 2026. Majestic listed 17 Nov 2026, relisted 17 Nov 2026. Link data listed by Majestic. Trust Flow, Citation Flow and Topical Trust Flow are trademarks of Majestic-12 Ltd.
phishclean-mcp-guard domain listing data
Show the listing data
| Domain age | 8 months |
|---|---|
| Registered | |
| Expires | |
| Registrar | NameCheap, Inc. |
| DNSSEC | Not signed |
phishclean-mcp-guard launch receipt entry
| Measure | At launch, then latest |
|---|---|
| Domain Rating (Ahrefs) | 32 to 32 no change since launch 17 Nov 2026, then 20 Nov 2026 |
| Trust Flow (Majestic) | 0 first measured 17 Nov 2026 |
| Referring domains (Majestic) | 8 first measured 17 Nov 2026 |
Day 2 since launch. Next receipt: day 30, 17 Dec 2026. We show what changed, not what caused it: many things affect these numbers. Nothing here is estimated.
Before you launch
The playbook on SaaS Directory List lists what to prepare, what to post on launch day and where else to list your product.
Compare and explore
Quick answers
What is the lethal trifecta for AI agents?
An agent that can read private data, reads content written by outsiders like web pages and emails, and can send data out. Together, hidden instructions in outside content can make the agent collect your data and send it away. The audit shows which of your MCP servers supply each part.
Is my MCP config sent to PhishClean?
No. The audit runs in your browser. Configs often contain tokens, so nothing you paste leaves the page. Tokens are shown masked in the results.
What is an MCP rug pull?
An MCP server that changes its tools after you approved it, usually through an update of a package your config runs without a pinned version. MCP Guard records each tool's definition the first time and blocks tools whose definition later changes until you accept the change.
Does MCP Guard send my data anywhere?
No. It runs on your machine as a proxy between the agent and each MCP server. Its log keeps tool names and argument names, not values. The only request it makes itself is to your own Slack or Discord webhook, if you set one.
Leave a review for phishclean-mcp-guard
No reviews listed for phishclean-mcp-guard yet. List the first one.
Listing a review is free. No one pays for reviews. See the listed reviews
List reviews · List updates · Share the list · Your tool? Claim this list entry · Report this list entry
phishclean-mcp-guard has a free listing. The owner can unlock the full profile with a paid launch for $5:
- Founder story (locked)
- Milestones (locked)
- Demo video (locked)
- Dofollow website link: not on this free listing (its website link is nofollow)
Also on the list
-
Number 1: CensysInternet intelligence platform for real-time threat detection and exposure management
CybersecurityUnited StatesSince 2017
DR 77 no change since last check 0 votes. Log in to move Censys up the list -
DR 76 +1 (up) since last check 0 votes. Log in to move Calif up the list
-
Number 3: WorldSignals
Free vulnerability scanner for exploited CVEs on your machines
CybersecurityEstonia
DR 19 no change since last check 0 votes. Log in to move WorldSignals up the list -
Number 4: KeybrumAutomated penetration testing for web apps that finds and proves every security leak
DR 6 +6 (up) since last check 0 votes. Log in to move Keybrum up the list