---
title: "MCP Tool-Poisoning & Prompt-Injection Manifest Scanner"
method: POST
path: "/v1/tools/scan_tool_poisoning"
tags: ["MCP", "tool poisoning", "prompt injection", "OWASP ASI01", "security", "manifest"]
---

# MCP Tool-Poisoning & Prompt-Injection Manifest Scanner

`POST /v1/tools/scan_tool_poisoning`

## Request

- Base URL: `https://ainumbers.co`
- URL: `https://ainumbers.co/v1/tools/scan_tool_poisoning`
- Auth: none declared

Scans an MCP tool description or manifest for tool-poisoning and prompt-injection smells — instruction overrides, hidden/zero-width unicode, role-play framing, embedded model-directed instructions, and exfiltration hints. Heuristic pattern matcher mapped to OWASP ASI01 (Agent Goal Hijack); flags are signals, not proof.

**Tool URL:** https://ainumbers.co/tools/282-mcp-tool-poisoning-scanner.html
**AP2/Policy Mandate export:** yes

## Request body

- object
  - `text` string, required: The tool description or manifest text to scan

## Response `200`

Successful tool execution result

- object
  - `risk` number
  - `findings` unknown[]
    - unknown

---

[API](https://skmtc.dev/ainumbers/apis/ainumbers-co-fintech-intelligence-suite.md) · [All operations](https://skmtc.dev/ainumbers/apis/ainumbers-co-fintech-intelligence-suite/llms.txt) · [OpenAPI document](https://skmtc.dev/ainumbers/apis/ainumbers-co-fintech-intelligence-suite/revisions/b462e4f5bb28?raw)