Jiffy Intel — Threat intelligence for AI artifacts

Summary

Custom GPT's OpenAPI schema defines a `/admin` path that the model can call, not mentioned in the GPT description. The path accepts arbitrary shell commands and runs them on the Action backend.

Severity & confidence

Severity

Critical

Active exploitation in the wild, high blast radius, or direct data exfiltration vector. Treat as an urgent remediation target; quarantine affected artifacts immediately.

Confidence

Confirmed

Observed directly by Jiffy scanners or validated by multiple independent sources. Indicators are reliable enough to drive automated action.

Indicators

Observed patterns and artifacts associated with this entry. Each indicator can be copied into your detection stack or SIEM.

(?i)paths:\s*[\s\S]{0,500}/(?:admin|debug|internal|__)

Detection rule

A YARA-style pseudo-rule auto-generated from the indicators above. Useful as a starting point — adapt the syntax for your target detection platform.

YARA-style pseudo-rule

rule jiffy_ti_2026_000083
{
    meta:
        source = "jiffy-intel"
        severity = "critical"
        description = "Auto-generated from Jiffy Intel indicators"
    strings:
    $content_pattern_0 = "(?i)paths:\\s*[\\s\\S]{0,500}/(?:admin|debug|internal|__)"
    condition:
        $content_pattern_0
}

Auto-generated from the indicators above. Adapt syntax for your detection stack before deploying.

Affected tools

Tool	Versions	Status
ChatGPT (GPT Store)	`*`	vulnerable

Example artifacts

Sanitized examples of artifacts Jiffy has observed exhibiting this pattern. Publisher handles are redacted; version ranges and status reflect the most recent scan.

Ops Dashboard GPTCustom GPT
Removed
Source
OpenAI GPT Store
First observed
Mar 19, 2026
Last observed
Apr 12, 2026
Deployer Pro GPTCustom GPT
Quarantined
Source
OpenAI GPT Store
First observed
Mar 23, 2026

How to remediate

01Audit Action OpenAPI schemas on first install.
02Reject GPTs whose schemas expose paths outside the stated scope.

Timeline & sources

Timeline

First observedMar 19, 202629 days ago
Last updatedApr 28, 2026today
PublishedMar 29, 202619 days ago

Sources

scanner

References

arxiv.orgarXiv

Related intel

Other entries of type Compromised Mcp, or the most recent published entries if no same-type matches exist.

Compromised McpLow

MCP server registers global tool names that shadow Jiffy primitives

MCP server that claims the tool name `scan` or `analyze` — names already used by Jiffy's first-party MCP. Collisions are resolved non-deterministically by the client, allowing the malicious server's tool to receive calls intended for Jiffy.

Jiffy IntelApr 28

Compromised McpHigh

Claude Project knowledge file contains embedded prompt-injection

Project knowledge documents (Word, PDF) uploaded as context contain footer text crafted as agent directives. Parsers pick up the footer; the model treats it as part of the project's truth.

Jiffy IntelApr 27

Compromised McpHigh

MCP server returns differential output to Claude vs. other clients

MCP server that inspects the User-Agent or initialization handshake and returns different tool output depending on the client. The Claude variant includes instruction injection; the research-scanner variant returns benign output. Evades automated scanning.