Google just moved security into the pull request. its pre-submit pipeline runs ai agents over every…
google just moved security into the pull request. its pre-submit pipeline runs ai agents over every infrastructure code change using a mantis multi-agent harness with local threat models, cutting false positives to 3 percent in some systems and proving the rest with real exploits through code-mender.
the security team is becoming the security harness.
Context
Verified source: Google's open source Mantis repository describes a toolkit of security review skills for AI coding agents to find, reproduce and patch vulnerabilities, and its README says all findings must be manually verified by a security expert. Google Cloud describes CodeMender as a security harness agent that finds and fixes code vulnerabilities.
Mantis and CodeMender exist as described in Google's pages. The pre-submit pipeline on every infrastructure code change, the 3 percent false positive rate and proving findings with real exploits were not seen in the excerpts read, so unsupported here, not refuted. Mantis's own README warns that models can hallucinate findings. 'The security team becomes the security harness' is the author's opinion.
Watch next
- Find Google's own write-up of the pre-submit pipeline and its false positive numbers.
Sources
- google/mantis, GitHubgithub.com
- Mantis README_AGENTS, GitHubgithub.com
- CodeMender: AI Agent for Code Security, Google Cloudcloud.google.com
Provenance
The note above is reproduced unedited from the original post, first published on Threads on 24 September 2026 at 14:50 IST. Sources are the papers and datasets the note draws on.
View the original post ↗Embed this note
More notes
The air is now being asked to keep its own ledger
the air is now being asked to keep its own ledger: ecmwf’s aifs compo becomes the first ai model to forecast atmospheric composition globally every three hours, cleanair simulates 365 days of pm2.5 over china in ten seconds, and a unified framework maps six pollutants at one kilometer across the whole country. the air now files its own composition report.
read the note →The current is now being asked to draw its own map
the current is now being asked to draw its own map: china’s langya 2.0 predicts six ocean phenomena including internal waves and mesoscale eddies, a deep net called wenhai resolves eddies globally with air sea flux formulas built in, and scripps infers surface currents from the way temperature patterns deform in satellite images. the ocean now files its own circulation report.
read the note →The soil is now being asked to report its own carbon
the soil is now being asked to report its own carbon: a nix color sensor paired with generative data augmentation predicts soil organic carbon without a lab, random forest drives 74 percent of soil health mapping studies, and sentinel 2 tracks five year carbon change across france and italy from 922 samples. the dirt now files its own carbon account.
read the note →