promptfoo Files

What's Changed

New Features

• feat(webui): First-class support for zooming eval results table by @will-holley in [#4966]
• feat: Grading and test generation improvements for BFLA, BOLA and RBAC by @sklein12 in [#4982]
• feat: New Sample Target by @sklein12 in [#4979]
• feat(webui): Apply metrics filter when clicking on a metric pill rendered in eval results cell by @will-holley in [#4991]
• feat: HTTP Target test button improvements by @faizanminhas in [#5007]
• feat: Add metadata filtering to eval results by @will-holley in [#5014]

Bug Fixes

• fix: add goal related rubric when grade crescendo turns to increase grading accuracy by @MrFlounder in [#4980]
• fix: update HTTP config generator endpoint to use v1 API by @mldangelo in [#4989]
• fix(redteam): find plugin assertion in strategy providers by @MrFlounder in [#4981]
• fix(site): dark mode style on redteam setup ui by @mldangelo in [#5000]
• fix: View logs button on redteam report by @sklein12 in [#5009]
• fix: undo unintended changes to http config editor by @faizanminhas in [#5012]
• fix: Autofocus on Redteam configuration description field by @sklein12 in [#5019]
• fix: remove filter icon by @sklein12 in [#5021]
• fix: Ollama token usage by @SamPatt in [#5022]
• fix(test): improve share test isolation to prevent CI timeouts by @mldangelo in [#5038]

Chores

• chore: revert eval view ui improvements by @mldangelo in [#4969]
• chore(webui): Improvements to pagination "go to" functionality by @will-holley in [#4976]
• chore(webui): Eval results sticky header improvements by @will-holley in [#4978]
• chore: update custom strategy prompt by @MrFlounder in [#4994]
• chore(cli): add support for 'help' argument to display command help by @mldangelo in [#4823]
• chore(examples): remove redteam-agent example by @mldangelo in [#5001]
• chore(providers): add GEMINI_API_KEY environment variable support by @mldangelo in [#5004]
• chore(webui): Migrate from JS to CSS for eval results scroll effects by @will-holley in [#4995]
• chore(webui): Eval result pagination UX improvements by @will-holley in [#4993]
• chore: Sort imports and turn on rule against unused imports by @faizanminhas in [#5010]
• chore: Make default target stateful by @faizanminhas in [#4992]
• chore: add medical plugins collection by @MrFlounder in [#5006]
• chore: Improve grading accuracy with Goal-Aware Grading for iterative/iterative tree by @MrFlounder in [#4996]
• chore: Add additionalRubric and storedGraderResult to GOAT and Custom providers by @MrFlounder in [#5015]
• chore: prevent testGenerationInstructions from being serialized if not present by @faizanminhas in [#5029]
• chore: Add lint rule to ensure key in jsx by @faizanminhas in [#5034]
• chore(webui): Eval Results UI Tweaks by @will-holley in [#5023]
• chore: skip goal extraction for datasets by @MrFlounder in [#5036]
• chore(providers): add GitHub Models provider by @mldangelo in [#4998]
• chore: bump version 0.117.2 by @MrFlounder in [#5045]
• ci: increase build job timeout from 4 to 5 minutes by @mldangelo in [#5043]
• test: refactor share.test.ts to prevent flaky timeouts by @mldangelo in [#5037]
• test: remove share.test.ts file by @mldangelo in [#5044]

Docs

• docs(providers): update OpenAI Assistants example by @aloisklink in [#4987]
• docs(redteam): improve custom strategy documentation by @mldangelo in [#4990]
• docs(blog): correct author attribution in DeepSeek censorship post by @mldangelo in [#5002]
• docs(openai): remove gpt-4.5-preview references after API deprecation by @mldangelo in [#5005]
• docs: remove label from featured blog post by @typpo in [#5011]
• docs(site): vegas contact redirect by @typpo in [#5033]
• docs(browser): improve browser provider documentation and examples by @mldangelo in [#5030]
• docs(providers): remove deprecated claude-3-sonnet-20240229 model references by @mldangelo in [#5018]
• docs(site): add hipaa badge by @typpo in [#5039]
• docs(site): add documentation for using text and embedding providers with Azure by @mldangelo in [#5027]
• docs(blog): fix missing blog posts by removing even-number enforcement by @mldangelo in [#5042]

Dependencies

• chore: bump @aws-sdk/client-bedrock-runtime from 3.846.0 to 3.848.0 by @dependabot in [#4985]
• chore: bump the npm_and_yarn group with 2 updates by @dependabot in [#4984]
• chore: bump @anthropic-ai/sdk from 0.56.0 to 0.57.0 by @dependabot in [#5016]
• chore: bump openai from 5.10.1 to 5.10.2 by @dependabot in [#5024]
• chore: bump the npm_and_yarn group with 2 updates by @dependabot in [#5026]
• chore: bump axios from 1.10.0 to 1.11.0 in the npm_and_yarn group by @dependabot in [#5031]

New Contributors

• @SamPatt made their first contribution in [#5022]

Full Changelog: https://github.com/promptfoo/promptfoo/compare/0.117.1...0.117.2