# Changelog

> **This is the PUBLIC form of this artifact. Findings are listed in full, but the details of SECURITY findings — which rule fired, in which file, on which line, and how to fix it — are deliberately withheld, and any secret-scanner results are excluded entirely. Where detail is absent here it was REMOVED FOR PUBLICATION; it is not missing from the analysis. The complete artifact is available from the repository owner.**

## Score

- CAI 70 → 69 (-0.9)
- Rubric changed (rubric-2026.10.5 → rubric-2026.11.4) — scores are not directly comparable.

## Lenses

- Code Health 80 → 79 (-1.2)
- Architecture 84 → 95 (+11.5)
- Maturity 67 → 73 (+6.2)
- Readiness 69 → 60 (-8.7)
- Security 70 → 76 (+5.6)

## Resolved (33)

- Duplicated block (18–31 lines × 2) (langextract/_compat/inference.py)
- Duplicated block (21 lines × 2) (langextract/providers/schemas/gemini.py)
- FileTooLong: scripts/create_provider_plugin.py (scripts/create_provider_plugin.py)
- High: security finding (details withheld)
- High: security finding (details withheld)
- High: security finding (details withheld)
- High: security finding (details withheld)
- High: security finding (details withheld)
- High: security finding (details withheld)
- High: security finding (details withheld)
- High: security finding (details withheld)
- High: security finding (details withheld)
- High: security finding (details withheld)
- High: security finding (details withheld)
- High: security finding (details withheld)
- High: security finding (details withheld)
- High: security finding (details withheld)
- High: security finding (details withheld)
- High: security finding (details withheld)
- High: security finding (details withheld)
- …and 13 more

## New (6)

- Edited copies of one member (3 members, 18+ corresponding lines) (langextract/_compat/inference.py)
- Edited copy of a member (27 corresponding lines) (langextract/providers/schemas/gemini.py)
- No ADRs found (pyproject.toml)
- OpenAILanguageModel._build_chat_completions_params (cognitive 16) (langextract/providers/openai.py)
- OpenAILanguageModel._build_chat_completions_params (cyclomatic 16) (langextract/providers/openai.py)
- Scattered collaborators

## Changes since last survey

- 13 commits — 12 feature/other, 1 fixes

## By area

- langextract/providers — 3 commits
- .github/workflows — 2 commits
- langextract/data_lib.py — 2 commits
- tests/test_live_api.py — 2 commits
- (root) — 1 commit
- examples/ollama — 1 commit
- langextract/core — 1 commit
- skills/langextract-usage — 1 commit

## Notable commits

- fix: fix: write Ollama visualization HTML as UTF-8 (#550)
- change: Harden pull_request_target workflows (#595)
- change: Load annotated documents saved with null extractions (#584)
- change: Move OpenAI live tests off retiring GPT-4o models (#601)
- change: Pin GitHub Actions to commit SHAs and harden workflow scripts (#606)
- change: Prepare v1.7.2 release (#607)
- change: Replace retired o4-mini in the live reasoning-effort test (#594)
- change: Retokenize a Document after its text changes (#583)
- change: Route gpt-6 model IDs to the OpenAI provider (#599)
- change: Send max_output_tokens as max_completion_tokens to OpenAI (#600)
- change: Stop dict_to_annotated_document from mutating its input (#577)
- change: Support httpx2 transport errors in Gemini retries (#605)
- change: Warn when model output stops at the token limit (#592)

## Architecture

- Unchanged — 0 containers · 1 contexts · 0 edges
