Commit History
-
Publish final submission package
Verified signatureCo-authored-by: Codex <codex@openai.com>502016b -
Add verdict confidence and trusted notes (#3)
- Add verdict confidence and trusted notes (71e8c6175e06c48e223044532cb90274c8851b26) - Serve kitchen UI only (80056c9f4b69e37a992592a5f72d41ac055937cb)dd109f9
-
Publish kitchen table UI
Verified signatureSync production kitchen table UI from GitHub commit 05e76bdadc9661e50d236aaaa7f6b1be52d6ec06. Co-authored-by: Codex <codex@openai.com>46412d9
-
Add heuristic guard test
Verified signatureCo-authored-by: Codex <codex@openai.com>95102dd -
Add guardrail tests
Verified signatureCo-authored-by: Codex <codex@openai.com>b5afd8a -
Stream scan progress states
Replace the static checking template with generator-driven staged progress before the MiniCPM call and keep the final report render unchanged. Co-authored-by: Codex <codex@openai.com>169919f -
Fix copy plan preview styling
Render the trusted-person copy plan as a light preview block, copy from a button data attribute, and humanize additional model tactic labels. Co-authored-by: Codex <codex@openai.com>71e0dc2 -
Polish v3 runtime output
Increase the default MiniCPM generation budget to avoid truncated JSON, humanize model-style labels in the report UI, and keep the copy-plan preview readable. Co-authored-by: Codex <codex@openai.com>1c0c50f -
Ship MiniCPM LoRA v3
Make the v3 MiniCPM LoRA adapter the default ZeroGPU model path, add v2/v3 hard-case data generation, record Modal training/eval evidence, and update submission docs. Co-authored-by: Codex <codex@openai.com>cc862d5
-
Add sanitized field scam examples
Co-authored-by: Codex <codex@openai.com>113d9bf -
Add Jawbreaker training and fallback spine
Co-authored-by: Codex <codex@openai.com>6fa99e6 -
Improve trusted handoff copy
Co-authored-by: Codex <codex@openai.com>358049d -
Add grandma-friendly Jawbreaker UI
Co-authored-by: Codex <codex@openai.com>aa9a872 -
Fix ZeroGPU scan memory flow
Co-authored-by: Codex <codex@openai.com>f97653f -
Add ZeroGPU transformers backend
Co-authored-by: Codex <codex@openai.com>bb6c547 -
Guard weak model outputs with scam heuristics
Co-authored-by: Codex <codex@openai.com>5311a14 -
Wire app to configurable model backend
Co-authored-by: Codex <codex@openai.com>8c85b65 -
Add model bakeoff evaluation harness
Co-authored-by: Codex <codex@openai.com>5971541 -
Build 100-case scam eval spine
Co-authored-by: Codex <codex@openai.com>292a298
-
Initialize Jawbreaker hackathon scaffold
6103e1f