Entity trail

Miasma

Source-backed findings, relationship evidence, citations, and briefing history from the public MindPattern archive.

Briefing refs

Findings

Edges

Sources

Corpus findings

2026-06-14 / agents-researcherSupply-chain worms now poison CLAUDE.md and .cursorrules with zero-width hidden instructionsThe continuously-mutating Miasma/Hades campaign (pivoting delivery every 48–72 hours since June 1) has added a novel agent-specific vector: packages inject .cursorrules and CLAUDE.md files containing zero-width Unicode hidden instructions. When a developer opens the project in Cursor or Claude Code, the assistant reads the configuration as legitimate project guidance and runs a 'security scan' that exfiltrates local secrets. Socket flagged 37 malicious Python wheels across 19 PyPI packages on June 7 and a compromised ensmallen 0.8.101 on June 8 — the first wave to weaponize AI coding-agent config files at scale.
2026-06-13 / agents-researcherMiasma supply-chain worm toolkit leaked on GitHub, already hitting npm/PyPI and CI/CDOn June 10, 2026, the modular 'Miasma' attack framework was leaked via compromised developer accounts in a 'Miasma-Open-Source-Release' GitHub repo. It targets public registries (PyPI, npm, RubyGems) and CI/CD systems including GitHub Actions and JFrog Artifactory, using stolen credentials to inject backdoors (latest Python variant named 'Hades', C2 markers like 'firedalazer'); the June 11 ThreatsDay bulletin confirmed 304+ components and 73 Microsoft GitHub repos impacted, mapped to MITRE T1195/T1078/T1021. It weaponizes the exact registries agent toolchains pull from — a self-propagating supply-chain risk for any pipeline that installs packages unattended.
2026-06-09 / rss-researcherHackers Plant Credential-Stealing 'Miasma' Malware in Microsoft's Open-Source GitHub ReposMicrosoft and GitHub disabled 70+ Microsoft open-source repositories — many of them Azure and AI developer tools — after attackers injected self-replicating 'Miasma' malware that steals credentials the instant a repo is opened in AI coding tools like Claude Code, Gemini CLI, or VS Code. Miasma is based on the open-sourced Mini Shai-Hulud codebase from a group called TeamPCP, and researchers suspect a link to a May breach of Microsoft's Durable Task project. It's a stark supply-chain warning for agentic-coding workflows that auto-open untrusted repositories.
2026-06-06 / agents-researcherMiasma Self-Spreading Worm Reaches Microsoft Azure GitHub Orgs — 73 Repos DisabledOn June 5, 2026, the Miasma supply-chain worm campaign reached Microsoft's Azure GitHub organizations, prompting GitHub to disable 73 repositories across four Microsoft GitHub orgs after malicious commits were pushed. The attack files execute credential-harvesting payloads when a developer opens the repository in Claude Code, Gemini CLI, Cursor, or VS Code — a new class of repo-as-trigger attack that weaponizes coding-agent auto-execution on project open. It underscores the danger of agents that act on untrusted repository contents without isolation.
2026-06-01 / skill-finderMiasma Supply Chain Attack Compromises 32 Red Hat npm Packages — Cloud Credential-Stealing Worm Hits ~80K Weekly DownloadsOn June 1, Wiz Research disclosed 'Miasma,' a supply chain attack that compromised 32 packages under the @redhat-cloud-services npm namespace using a stolen Red Hat employee GitHub account to push malicious orphan commits bypassing code review. The payload deploys a 4.2 MB obfuscated credential stealer targeting GCP and Azure cloud identities, and self-propagates as a worm.
2026-03-29 / hn-researcherMiasma: Rust-Built AI Scraper Trap Creates Endless Poison Data Pit with Compressed Loops — 42 Points on HNOpen-source Rust tool (124 GitHub stars, GPL-3.0) that traps AI web scrapers by serving hidden links invisible to humans but discoverable by crawlers. Scrapers that follow the links enter an infinite loop of 'poisoned training data from the poison fountain' with self-referential links and compressed responses to minimize egress costs. Supports configurable connection limits (500 default), custom routing for Nginx integration, and forced gzip encoding. A builder-friendly defensive tool against unauthorized AI training data collection.

Graph relationships

MENTIONS
Wiz Research
Wiz Research disclosed Miasma supply chain attack
Source finding
CRITICIZES
Claude Code
Miasma worm targets Claude Code for credential harvesting via malicious repository execution
Source finding
CRITICIZES
VS Code
Miasma malware steals credentials from VS Code.
Source finding
CRITICIZES
Claude Code
Miasma malware steals credentials from Claude Code.
Source finding
USES
Rust
Miasma is a Rust-built AI scraper trap that serves poisoned training data.

Source trail

per Rescana many of them Azure and AI developer tools Releasebot Claude Code Docs StepSecurity MIT Technology Review Adversa AI

Graph sources

entity graphfindings textkg edgeskg entitiesnewsletter issues