Loading vulnerability…
CVE-2026-41679 is a critical missing-authorization vulnerability in Paperclip versions prior to 2026.416.0. In the default network-accessible authenticated deployment, an unauthenticated remote attacker could self-register without email verification, create and self-approve a CLI authorization challenge, and obtain a persistent board-level API credential. Although direct creation of a company required instance-administrator privileges, the new-company import workflow incorrectly accepted board-level access. An attacker could import an agent configuration using Paperclip's process adapter and activate the agent, causing an attacker-selected command to execute as a child process of the Paperclip server.
Mallory correlates every CVE against your assets, your vendors, and active adversary campaigns. Know which vulnerabilities matter for you, not just which ones are loud.
What it means. What to do now. Patch path, mitigations, and the assume-compromise checklist.
What an attacker gets, and what they’ve been doing with it.
If you can’t patch tonight, do this now.
Patch, then assume compromise.
1 valid exploit after Mallory filtered fakes, detection scripts, and README-only repos (1 hidden).
This repository is a small Python proof-of-concept exploit for CVE-2026-41679 affecting Paperclip AI. The repo contains one executable script (poc.py), a short README with usage instructions, and a minimal requirements.txt listing requests. The exploit is not framework-based. The script performs a full exploitation chain against a target Paperclip instance over HTTP(S): it signs up a user, signs in to obtain a session cookie, creates a CLI auth challenge, approves that challenge with the authenticated session, then uses the returned board API token to call the company import API. In that import request it supplies inline files including COMPANY.md, agents/pwn/AGENTS.md, and a malicious .paperclip.yaml. The YAML defines an agent named pwn using a process adapter configured to run bash with '-c' and attacker-controlled commands. After import, the script calls the agent wakeup endpoint to trigger execution. The exploit’s main capability is authenticated remote command execution through abuse of the import and agent execution workflow. The default payload writes the output of 'id' and 'whoami' to /tmp/pwned.txt, but the -c argument allows arbitrary shell commands, making this more than a pure detector. TLS certificate verification is explicitly disabled, so it will also work against instances with self-signed certificates. Overall, this is an operational PoC that automates exploitation end-to-end and demonstrates code execution on the target host.
Products and vendors Mallory has correlated with this vulnerability. Open in Mallory to drill down to specific CPE configurations and version ranges.
Vendor-confirmed product mapping. Mallory continuously reconciles this list against your asset inventory.
24 sources tracked across advisories, community write-ups, and news. New activity surfaces here as Mallory finds it.
A vulnerability identified as CVE-2026-41679 is referenced solely in the creation of a YAML file; the content provides no technical details, affected product, impact, or exploitation information.
A critical authorization bypass vulnerability in the Paperclip AI management platform that can be chained with privilege escalation and malicious agent configuration import to achieve arbitrary code execution and full instance takeover.
A critical authorization bypass in Paperclip that lets a lower-privileged user import a malicious company bundle and achieve remote command execution on the host via agent configuration and the built-in process adapter.
A critical authorization bypass in the Paperclip AI management platform that could let a remote attacker self-register, gain board-level API access, and achieve arbitrary code execution with the permissions of the Paperclip server process.
Query your assets running an affected version, and investigate the blast radius.
Every observed campaign linking this CVE to a named adversary.
Malware families riding this exploit, with evidence and IOCs.
YARA, Sigma, Snort, and vendor rules, auto-deployed to your SIEM.
Cross-references every affected SKU, including bundled OEM variants.
Community discussion across Reddit, Mastodon, and other social sources.