VQV Signal

BIG MOVE WATCH TECHNICAL

Microsoft's agent-framework PyPI package hijacked to steal cloud and CI/CD credentials

According to a post on X by @AikidoSecurity, versions 1.26.1 and 1.26.2 of agent-framework, the Python package for Microsoft Agent Framework, carried a startup hook that sends environment variables and Azure, GitHub Actions and other CI/CD credentials to an attacker. The package has about 190,000 downloads a month. Aikido says anyone who installed it between about 12:35 and 1:20 PM PT is affected, including users who pulled it in through a loose ">=1.26" requirement.

Source: Scobleizer News Pulse (X) · x.com Found via Robert Scoble’s Scobleizer News Pulse · a post on X, not independently verified Published 2026-11-18T00:20:00+00:00 Detected 2026-11-18T06:03:59+00:00
View original source

According to a post on X by @AikidoSecurity, versions 1.26.1 and 1.26.2 of agent-framework, the Python package for Microsoft Agent Framework, carried a startup hook that sends environment variables and Azure, GitHub Actions and other CI/CD credentials to an attacker. The package has about 190,000 downloads a month. Aikido says anyone who installed it between about 12:35 and 1:20 PM PT is affected, including users who pulled it in through a loose ">=1.26" requirement.

AI-assisted summary based on the listed source.

Versions 1.26.1 and 1.26.2 of agent-framework, the Python package for Microsoft Agent Framework (about 190,000 downloads a month), carried a startup hook that sends environment variables and Azure, GitHub Actions and other CI/CD credentials to an attacker; Aikido says anyone who installed between about 12:35 and...

Developers and teams that install agent-framework, especially in CI/CD pipelines, may have exposed cloud and repository credentials if they pulled one of the affected versions during that window. Anyone with a loose version requirement should check which version actually got installed. Watch for further details from Aikido and any official guidance from Microsoft or PyPI on removal and credential rotation.

Signal Strength 80% Technical label WATCH Public Interest 41 Category BIG MOVE Reader Depth TECHNICAL

Signal Strength reflects source quality, relevance, freshness and evidence. Public Interest helps organize discovery; it is not proof of truth.

Public Interest components
Recognizable Entity Score 51 Practical Impact Score 8 Novelty Interest Score 70 Consequence Score 38 Curiosity Score 16 Shareability Score 53

VQV surfaced this signal because it is recent, relevant to Developer Tools, connected to Scobleizer News Pulse (X).