RepoJournal
AI Red Teaming AI Red Teaming
29 wires and counting

$ follow AI Red Teaming

Keep up with AI Red Teaming in about 3 minutes: what actually shipped — the commits, pull requests, releases, and security advisories that matter.

or

fair warning: these emails are deeply technical. diffs, version numbers, CVEs, benchmark deltas. if that's not your idea of a good read, this isn't your newsletter.

Folds into your digest — weekly by default, monthly if you prefer. Unsubscribe in one click.

$ status

wire 2026-10-30
stories 26
RSS Archive Reviews

© 2026 RepoJournal Home Showcase How it works Privacy

$ the-wire · showcase

PyRIT tidies its warning UI and pins down converter state across prompt restores

By RepoJournal · Filed · About AI Red Teaming · Composed from the cited sources · methodology

richlundeen WatchTree-19 hannahwestra25
3 people shipped this
richlundeen @richlundeen 3 cited WatchTree-19 @WatchTree-19 1 cited hannahwestra25 @hannahwestra25 1 cited

PyRIT's day is mostly housekeeping for the people clicking through it: setup warnings get consolidated into one place, and the converter registry stops losing track of what you built.

FIX: Unify operator and operation setup warnings microsoft/PyRIT

by richlundeen

Operator and operation setup warnings now sit together under the warning triangle, with a subtle field highlight replacing the separate operation banner, and recovery controls stay where they were. Fewer places to look when a run refuses to start.

FEAT: Save temporary converters to the registry microsoft/PyRIT

by richlundeen

Temporary converters are now saved to the registry. If you build a converter ad hoc for a single run, it persists as a named entry rather than vanishing when you move on.

FIX: Preserve current converters when restoring sent prompts microsoft/PyRIT

by richlundeen

Restoring a previously sent prompt now preserves the converters currently configured instead of clobbering them. Re-running an old prompt keeps the setup you have loaded right now.

FEAT Legacy scorer hook errors and explicit objective conditions (phase 12, part 1) microsoft/PyRIT

by WatchTree-19

Groundwork for legacy scorer hook errors and explicit objective conditions, landing as part one of a multi-phase effort. Nothing an operator can use yet; the pull request text itself carries an unresolved note about breaking changes.

FIX CoPyRIT deployment job overrides and build disk cleanup microsoft/PyRIT

by hannahwestra25

The CoPyRIT deployment job overrides and build disk cleanup changed, which keeps the shipped deployment path building rather than altering anything you call from PyRIT itself.

$ ls ai-red-teaming/reviews/ # the step-back reads

the week in review promptfoo 0.125.0 makes Claude and Codex SDKs opt-in
Full archive → Weekly reviews →

Quick answers

What shipped in AI Red Teaming on October 30, 2026?
PyRIT's day is mostly housekeeping for the people clicking through it: setup warnings get consolidated into one place, and the converter registry stops losing track of what you built. In total, 12 commits and 14 pull requests landed.
Who contributed to AI Red Teaming on October 30, 2026?
3 developers shipped this update, including richlundeen, WatchTree-19, and hannahwestra25.
What were the notable AI Red Teaming updates?
FIX: Unify operator and operation setup warnings, FEAT: Save temporary converters to the registry, and FIX: Preserve current converters when restoring sent prompts.
♥

$ log AI Red Teaming --recent

Anthropic SDK tracing collides with promptfoo spans, and PyRIT adds a new attack technique 37 shipped PyRIT adds tool role-play attacks, garak moves probe intents into reusable stubs 24 shipped PyRIT validates penalty parameters, corrects GCG gradient aggregation 39 shipped Promptfoo moves off vulnerable MSAL, PyRIT adds a char-split attack and cloud token detection 69 shipped

$ log --all # elsewhere on the wire

Alignment & Post-training XPU rollout images, a tensordict 0.11 data fix, and a gated-judge guard Agent Security Denied RAG queries get real hashes, Docker deployer logs gain stderr

$ ls showcases/

Safety Evals Agent Security Interpretability Alignment & Post-training

$ init your-desk

RepoJournal can write one of these for any repo you care about. Sign in with GitHub, pick your repos and sources, and your desk gets a wire like this one every morning.