gcdRadar
archive/Research/AIR-2026-0345
closedclosed 20 Feb 2026Research1 Research / proof of conceptAI

Poisoned ClawdBot Skill Distributed on Registry

ClawdBot, ClawdHub

A researcher demonstrated a supply chain attack using a poisoned skill on a registry that triggered prompt injection and shell command execution. Multiple users downloaded the compromised package during the test.

first report
6 Feb 2026
last update
6 Feb 2026

Timeline · 2

  1. 6 Feb 2026Supply Chain Compromise via Poisoned ClawdBot Skill
  2. 6 Feb 2026eating lobster souls Part II: the supply chain (aka - backdooring the #1 downloaded clawdhub skill)

Sources · 2

News and write-ups

  1. 2026-02-06

Research and catalogues

  1. 2026-02-06
    Supply Chain Compromise via Poisoned ClawdBot Skill
    MITRE ATLAS · atlas.mitre.org/studies/AML.CS0049

Titles and links belong to their publishers. The summary above is written by gcdRadar in its own words; data is CC BY-SA 4.0.

Classification

AI role
Attack on AI
categories
Supply chainPrompt injectionAgent hijack
severity
1 · Research / proof of concept
targets
Framework / tooling
dates
happened 26 Jan 2026 · first reported 6 Feb 2026
origin
MITRE ATLAS
Derived from MITRE ATLAS (Apache 2.0, © MITRE).
record
confidence 0.95 · created 2026-10-18 06:55Z · updated 2026-10-18 06:55Z

Lane, AI role, categories and severity come from the classifier; the OWASP and ATLAS rows are derived from the categories. How these are assigned

Copy a brief

This record as a short Markdown brief, for a ticket, a chat or a coding agent.

the brief as text
# Poisoned ClawdBot Skill Distributed on Registry

https://gcdradar.vercel.app/incidents/poisoned-clawdbot-skill-distributed-on-registry · AIR-2026-0345 · closed 2026-02-20

- Lane: Research
- Severity: 1 (Research / proof of concept)
- Fix status: Fix status unknown
- First reported: 2026-02-06 (happened 2026-01-26)
- AI role: Attack on AI
- Products: ClawdBot, ClawdHub

## Summary

A researcher demonstrated a supply chain attack using a poisoned skill on a registry that triggered prompt injection and shell command execution. Multiple users downloaded the compromised package during the test.

## What to do

No action recorded; read the sources below and the vendor's advisory.

## Latest updates

- 2026-02-06: eating lobster souls Part II: the supply chain (aka - backdooring the #1 downloaded clawdhub skill) (https://x.com/theonejvo/status/2015892980851474595)
- 2026-02-06: Supply Chain Compromise via Poisoned ClawdBot Skill (https://atlas.mitre.org/studies/AML.CS0049)

## Sources

- 2026-02-06 MITRE ATLAS: Supply Chain Compromise via Poisoned ClawdBot Skill https://atlas.mitre.org/studies/AML.CS0049
- 2026-02-06 x.com: eating lobster souls Part II: the supply chain (aka - backdooring the #1 downloaded clawdhub skill) https://x.com/theonejvo/status/2015892980851474595

Verify against the vendor's advisory before changing production. Data CC BY-SA 4.0, https://gcdradar.vercel.app.

Related

  1. closed9.8
    Langflow CVE-2026-27966: Langflow Remote Code Execution in CSV Agent

    Patch now · Disclosed · since 27 Feb 2026 · updated 27 Feb 2026 · CVE-2026-27966

    AI
  2. closed9.9
    Semantic Kernel RCE Vulnerability in Microsoft Semantic Kernel Search Plugin

    Research · since 7 May 2026 · updated 30 Jun 2026 · CVE-2026-26030

    AI
  3. closed
    AI
  4. closed
    AI
  5. closed
    AI
  6. closed
    AI