MCP Invisible-Unicode Tool Poisoning (Approval-View Fidelity Gap)

mcp-invisible-unicode-poisoning is a high severity check in the ai category, written in python. Its source is templates/ai/mcp/mcp-invisible-unicode-poisoning.py in cert-x-gen-templates.

What it checks

Detects MCP servers whose advertised tool, prompt, resource or instructions text carries Unicode that a human approval view never renders but the connected model still reads - TAG-block payloads, zero-width carrier runs, and Trojan-Source bidi controls

There is a playbook for this check

Precise MCP metadata scanning sets out the use case, the probe flow, the oracle, and the fixture that proves it in both directions.

Header

Field	Value
Id	mcp-invisible-unicode-poisoning
Severity	high
Language	python
Category	ai
Author	Bugb Research
Template version	1.0.0
Confidence	100
CVSS	8.2
Weakness	CWE-1427, CWE-1007
Tags	mcp, ai, agent, tool-poisoning, prompt-injection, unicode, trojan-source, cwe-1427
Target kind	http
Oracle	property

Conventions of the repository, not fields the engine reads

@target_kinds and @oracles are documented in the templates repository’s contribution guide, and cxg’s annotation parser does not carry them into a template’s metadata. Nothing at scan time selects on them. The annotation header schema is every field the parser does read.

References

Declared in the header. cxg parses @references and then discards it, and nothing at scan time reads it, so this is the only place the links a template cites are surfaced.

• https://arxiv.org/pdf/2607.05744
• https://modelcontextprotocol.io/specification
• https://cwe.mitre.org/data/definitions/1427.html
• https://cwe.mitre.org/data/definitions/1007.html
• https://trojansource.codes/

Run it

To see what the copy on your machine says about itself, and to confirm it is installed at all:

cxg template info mcp-invisible-unicode-poisoning

The id it prints is the one to pass anywhere a template is selected. See cxg template for the rest of the subcommand, Scan a target for running a scan, and A match is not a finding for how to read what comes back.

Related

• Template catalog is the whole corpus.
• ai is the rest of this category.
• Templates by weakness class groups checks by the defect they look for.
• Contribute a template is the route from a check you needed and could not find to a merged one.