v4.2.14

What's Changed

Commits since v4.2.13

f3f6fc8 Build dist for v4.2.14
d5463b7 Bump version to v4.2.14
cade021 add workflow_dispatch support for full PR review
0d5c238 auto-release patch versions on push to main

Full Changelog: v4.2.13...v4.2.14

What's Changed

• Additional Rules by @lightninglu10 in #43
• existing comments by @lightninglu10 in #44
• even more agentic by @lightninglu10 in #45
• updating package versions by @lightninglu10 in #46
• Creating better context by @lightninglu10 in #47
• do not fail on bad calls by @lightninglu10 in #48
• Tool finish by @lightninglu10 in #49
• Get back to output json completion by @lightninglu10 in #50
• Improve re-review and completion prompting by @lightninglu10 in #51
• Improve rereview prompting and add force full review by @lightninglu10 in #52
• test using gpt-latest instead of hardcoded model by @lightninglu10 in #53

Full Changelog: v4.0.10...v4.2.14

v4.2.13

What's Changed

Commits since v4.2.12

791409b Build dist for v4.2.13
babc9b2 Bump version to v4.2.13
02187af add real-world API model selection tests
6621a94 fix version extraction for Anthropic and Gemini models
791bf92 fix GPT model selection to prefer semantic version over date suffix
609d7d6 fix model alias patterns to match versioned suffixes (e.g., gpt-5.2-preview)
c3d1f1d fix model alias patterns to match decimal versions (e.g., gpt-5.2)
1ecdbbf add beta release support with auto-incrementing version

Full Changelog: v4.2.12...v4.2.13

What's Changed

• Additional Rules by @lightninglu10 in #43
• existing comments by @lightninglu10 in #44
• even more agentic by @lightninglu10 in #45
• updating package versions by @lightninglu10 in #46
• Creating better context by @lightninglu10 in #47
• do not fail on bad calls by @lightninglu10 in #48
• Tool finish by @lightninglu10 in #49
• Get back to output json completion by @lightninglu10 in #50
• Improve re-review and completion prompting by @lightninglu10 in #51
• Improve rereview prompting and add force full review by @lightninglu10 in #52
• test using gpt-latest instead of hardcoded model by @lightninglu10 in #53

Full Changelog: v4.0.10...v4.2.13

v4.2.12

What's Changed

Commits since v4.2.11

796f0b2 Build dist for v4.2.12
ba4a4f5 Bump version to v4.2.12
218dd3e Merge pull request #53 from quantfive/test-gpt-latest
098ff0e test using gpt-latest instead of hardcoded model
ccd0247 update all examples to use -latest aliases by default
e9b7739 expand model aliases documentation with all providers

Full Changelog: v4.2.11...v4.2.12

What's Changed

• Additional Rules by @lightninglu10 in #43
• existing comments by @lightninglu10 in #44
• even more agentic by @lightninglu10 in #45
• updating package versions by @lightninglu10 in #46
• Creating better context by @lightninglu10 in #47
• do not fail on bad calls by @lightninglu10 in #48
• Tool finish by @lightninglu10 in #49
• Get back to output json completion by @lightninglu10 in #50
• Improve re-review and completion prompting by @lightninglu10 in #51
• Improve rereview prompting and add force full review by @lightninglu10 in #52
• test using gpt-latest instead of hardcoded model by @lightninglu10 in #53

Full Changelog: v4.0.10...v4.2.12

v4.2.11

What's Changed

Commits since v4.2.10

446b1f1 Build dist for v4.2.11
ca5cf7b Bump version to v4.2.11
1e75a3a update xAI fallback to grok-4
f34fddd skip dynamic resolution for Mistral - they maintain their own -latest aliases
99327b3 add dynamic model resolution for all supported providers
588a4bd add dynamic model resolution for Google Gemini API
9a2c688 implement dynamic model alias resolution via provider APIs
72d5202 update model aliases to use provider-native auto-updating aliases where available
42c625f add "latest" model aliases for automatic version resolution
efcd41e handle force-pushed branches in re-review: fall back to full review
374a19c add workflow_dispatch example with force_full_review input
af63bad add force_full_review to README inputs table

Full Changelog: v4.2.10...v4.2.11

What's Changed

• Additional Rules by @lightninglu10 in #43
• existing comments by @lightninglu10 in #44
• even more agentic by @lightninglu10 in #45
• updating package versions by @lightninglu10 in #46
• Creating better context by @lightninglu10 in #47
• do not fail on bad calls by @lightninglu10 in #48
• Tool finish by @lightninglu10 in #49
• Get back to output json completion by @lightninglu10 in #50
• Improve re-review and completion prompting by @lightninglu10 in #51
• Improve rereview prompting and add force full review by @lightninglu10 in #52

Full Changelog: v4.0.10...v4.2.11

v4.2.10

What's Changed

Commits since v4.2.9

bf3cd0e Build dist for v4.2.10
0dfa90c Bump version to v4.2.10
bd380ef Merge pull request #52 from quantfive/improve-rereview-prompting
06b0386 add force_full_review option to force complete review of all files
842ca81 handle edge case: no file changes since last review
38436f2 clarify re-review scope: diff from last review commit, not just "new commits"
da7be5e fix initial message workflow to respect re-review scoping
dca6ce0 fix workflow order: determine scope BEFORE creating todos
febe328 add fallback options for getting diff in shallow clones

Full Changelog: v4.2.9...v4.2.10

What's Changed

• Additional Rules by @lightninglu10 in #43
• existing comments by @lightninglu10 in #44
• even more agentic by @lightninglu10 in #45
• updating package versions by @lightninglu10 in #46
• Creating better context by @lightninglu10 in #47
• do not fail on bad calls by @lightninglu10 in #48
• Tool finish by @lightninglu10 in #49
• Get back to output json completion by @lightninglu10 in #50
• Improve re-review and completion prompting by @lightninglu10 in #51
• Improve rereview prompting and add force full review by @lightninglu10 in #52

Full Changelog: v4.0.10...v4.2.10

v4.2.9

What's Changed

Commits since v4.2.8

43c10fd Build dist for v4.2.9
4918cdd Bump version to v4.2.9
3978246 Merge pull request #51 from quantfive/improve-rereview-prompting
30d7855 improve re-review and completion prompting

Full Changelog: v4.2.8...v4.2.9

What's Changed

• Additional Rules by @lightninglu10 in #43
• existing comments by @lightninglu10 in #44
• even more agentic by @lightninglu10 in #45
• updating package versions by @lightninglu10 in #46
• Creating better context by @lightninglu10 in #47
• do not fail on bad calls by @lightninglu10 in #48
• Tool finish by @lightninglu10 in #49
• Get back to output json completion by @lightninglu10 in #50
• Improve re-review and completion prompting by @lightninglu10 in #51

Full Changelog: v4.0.10...v4.2.9

v4.2.8

What's Changed

Commits since v4.2.7

8158af7 Build dist for v4.2.8
ea96a62 Bump version to v4.2.8
1a5c577 Merge pull request #50 from quantfive/output-json-complete
08d69d2 better prompting
28492d8 get back to output json completino

Full Changelog: v4.2.7...v4.2.8

What's Changed

• Additional Rules by @lightninglu10 in #43
• existing comments by @lightninglu10 in #44
• even more agentic by @lightninglu10 in #45
• updating package versions by @lightninglu10 in #46
• Creating better context by @lightninglu10 in #47
• do not fail on bad calls by @lightninglu10 in #48
• Tool finish by @lightninglu10 in #49
• Get back to output json completion by @lightninglu10 in #50

Full Changelog: v4.0.10...v4.2.8

v4.2.7

What's Changed

Commits since v4.2.6

93d760e Build dist for v4.2.7
8bdcb13 Bump version to v4.2.7
5bb17ec Merge pull request #49 from quantfive/tool-finish
f5d1c09 unlimited turns by default
4dbfaeb tool call to finish agent loop
94fe4d1 tool call to finish agent loop

Full Changelog: v4.2.6...v4.2.7

What's Changed

• Additional Rules by @lightninglu10 in #43
• existing comments by @lightninglu10 in #44
• even more agentic by @lightninglu10 in #45
• updating package versions by @lightninglu10 in #46
• Creating better context by @lightninglu10 in #47
• do not fail on bad calls by @lightninglu10 in #48
• Tool finish by @lightninglu10 in #49

Full Changelog: v4.0.10...v4.2.7

v4.2.6

What's Changed

Commits since v4.2.5

734e410 Build dist for v4.2.6
455b36a Bump version to v4.2.6
f350cfd Merge pull request #48 from quantfive/improved-prompting
58c277f do not fail on bad calls

Full Changelog: v4.2.5...v4.2.6

What's Changed

• Additional Rules by @lightninglu10 in #43
• existing comments by @lightninglu10 in #44
• even more agentic by @lightninglu10 in #45
• updating package versions by @lightninglu10 in #46
• Creating better context by @lightninglu10 in #47
• do not fail on bad calls by @lightninglu10 in #48

Full Changelog: v4.0.10...v4.2.6

v4.2.5

What's Changed

Commits since v4.2.4

45a349b Build dist for v4.2.5
18f6aaf Bump version to v4.2.5
8f8e1f1 Merge pull request #47 from quantfive/better-context
b119d4b creating better context

Full Changelog: v4.2.4...v4.2.5

What's Changed

• Additional Rules by @lightninglu10 in #43
• existing comments by @lightninglu10 in #44
• even more agentic by @lightninglu10 in #45
• updating package versions by @lightninglu10 in #46
• Creating better context by @lightninglu10 in #47

Full Changelog: v4.0.10...v4.2.5