---
id: "MAL-2026-15810"
kind: "package"
aliases: ["MAL-2026-15810"]
title: "Malicious code in gcphelpit (PyPI)"
severity: "critical"
cvss: null
epss: null
exploited: false
status: "open"
category: "package"
products: [{"name": "gcphelpit", "fixed": [], "affected": "= 0.1.0; = 0.1.1; = 0.1.2", "ecosystem": "pypi"}]
ecosystems: ["pypi"]
tags: ["malicious", "package"]
not_affected_if: []
published_at: "2026-09-01T16:19:02+00:00"
first_seen_at: "2026-10-26T20:51:56+00:00"
updated_at: "2026-10-26T21:12:36+00:00"
sources: [{"source": "osv-malicious", "url": "https://osv.dev/vulnerability/MAL-2026-15810", "first_seen_at": "2026-10-26T20:51:56+00:00"}]
---

# Malicious code in gcphelpit (PyPI)

## Affected

- pypi:gcphelpit affected = 0.1.0; = 0.1.1; = 0.1.2 fixed ?

## Timeline

- 2026-10-26T20:51:56+00:00 osv-malicious https://osv.dev/vulnerability/MAL-2026-15810

## From the sources

---
_-= Per source details. Do not edit below this line.=-_

## Source: amazon-inspector (6be1770a21667861e10f778ca485b270170e20a35a529bc2a041bc72266bdb7d)
The module src/gcphelpit/_verify.py runs at import time: it reads the installer's ~/.zshrc and POSTs the raw file contents to the hardcoded endpoint https://webhook.site/bc67d797-61ff-4f3b-a9b8-5df6a5e7643d. cli.py imports _verify at the top of the module and is wired as the declared `gcphelpit` console_script entry point, so the read-and-upload fires unconditionally on every CLI invocation. ~/.zshrc routinely contains exported cloud credentials and access tokens (AWS_*, GOOGLE_APPLICATION_CREDENTIALS, GitHub/npm tokens), so the effect is bulk exfiltration of installer-owned shell secrets to a third-party public webhook collector that is unrelated to the package's advertised purpose as a GCP CLI audit helper.

## Source: kam193 (d3e108475330381be537963456cb012b943f2d0a3693c83205f8f5b01f36635a)
During initialization of the CLI, the package exfiltrates sensitive files. Prior version 0.1.2 the code was launching a calculator as PoC instead of exfiltrating data.


---

Category: MALICIOUS - The campaign has clearly malicious intent, like infostealers.


Campaign: 2026-09-gcphelpit


Reasons (based on the campaign):


 - files-exfiltration

---

_Free to read at a human pace. An API with keys and an MCP server are in preparation; early access: hello@infrastack.ai_