* docs: add benchmark reproduction tooling and update benchmarks page

Add the benchmark comparison scripts (compare.sh, bench CLI source,
chart generator) to docs/about/benchmark-tools/ so anyone can reproduce
the GoModel vs LiteLLM benchmark. Update the benchmarks.mdx Mintlify
page with a "Reproduce it yourself" section including prerequisites,
quick start, and tuning instructions.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* fix: update Go version prerequisite to 1.25+

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

---------

Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>

* docs: add benchmark reproduction tooling and update benchmarks page

Add the benchmark comparison scripts (compare.sh, bench CLI source,
chart generator) to docs/about/benchmark-tools/ so anyone can reproduce
the GoModel vs LiteLLM benchmark. Update the benchmarks.mdx Mintlify
page with a "Reproduce it yourself" section including prerequisites,
quick start, and tuning instructions.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* fix: update Go version prerequisite to 1.25+

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* fix: use absolute GitHub URLs for benchmark-tools links in docs

Relative links (./benchmark-tools/) resolve within Mintlify instead of
pointing to the actual source files on GitHub. Replace with absolute
GitHub URLs and fix repository URL casing.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

---------

Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>

* docs: improve README SEO with keyword-rich headers and badges

Add Docker Pulls and Go Version badges, rename section headers to
include target keywords (AI Gateway, OpenAI-Compatible, LLM Providers,
Gateway Configuration), and add a Community section.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* docs: replace em dashes with hyphens in README

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* docs: fix Go Version badge link to use absolute GitHub URL

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

---------

Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>

* docs: add observability flags to Quick Start docker command

Enable logging, body/header capture, and text log format by default
in the primary Quick Start example so new users get full observability
out of the box.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* docs: add observability flags to quickstart.mdx docker command

Add the same LOGGING_ENABLED, LOGGING_LOG_BODIES, LOG_FORMAT, and
LOGGING_LOG_HEADERS flags to the docs quickstart page, matching the
README update.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* docs: add LOG_FORMAT to .env.template

Document the LOG_FORMAT environment variable in the server
configuration section of .env.template.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

---------

Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>

* feat: add GitHub-style contribution calendar heatmap to dashboard

Add a year-long activity heatmap to the Overview page showing daily
token consumption or cost data. Includes switchable Tokens/Costs modes,
hover tooltips, dark/light theme support, and responsive layout.

Backend: add cost fields (InputCost, OutputCost, TotalCost) to the
DailyUsage struct and update SQLite, PostgreSQL, and MongoDB readers
to aggregate and return cost data alongside token counts.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* Fix dashboard calendar refresh and cost aggregate nullability

---------

Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>

…ni reasoning param (#145)

* fix: surface Claude thinking blocks as reasoning_content and translate Gemini reasoning param

Claude's extended thinking content was silently dropped during Anthropic→OpenAI
normalization. Thinking blocks now flow through as `reasoning_content` in both
non-streaming and streaming responses, matching the format OpenAI SDKs expect.

Gemini's OpenAI-compatible endpoint requires `reasoning_effort` as a top-level
string rather than the nested `reasoning.effort` object — requests now translate
the parameter automatically, fixing 400 errors.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* fix: update extended-thinking golden file to include reasoning_content

The contract replay test golden file was stale — now includes the
reasoning_content field that the gateway correctly surfaces from
Anthropic thinking blocks.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

---------

Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>

* refactor: use different naming

* refactor: adjust naming

* chore: clean up

#137 (review)

* Update CLAUDE.md

Co-authored-by: coderabbitai[bot] <136622811+coderabbitai[bot]@users.noreply.github.com>

* chore: clean up

* Update CLAUDE.md

Co-authored-by: coderabbitai[bot] <136622811+coderabbitai[bot]@users.noreply.github.com>

---------

Co-authored-by: coderabbitai[bot] <136622811+coderabbitai[bot]@users.noreply.github.com>

* chore(release): rebuild automatic release notes pipeline

* docs: fix MD041 lint — promote PR template heading to H1

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

---------

Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>

* feat: add model aliases feature

* fix: use typed mongo alias id filters

* fix: return consistent aliases unavailable errors

* fix: classify alias validation and storage errors

* Fix registry lookup metadata race

* fix: preserve provider routing for aliases

* refactor: require file provider inventory

* Add inline alias management and early alias resolution

* Log alias usage in audit entries

* Fix alias resolution regressions

* Rewrite file-backed batch sources before submission

* Harden batch rewrite helpers

* Fix rewritten batch input file lifecycle

* Refine alias controls in models dashboard

* Fix disabled alias validation gaps

* Preserve sparse batch refresh state

* refactor: introduced execution plan and decoupling

* refactor gateway execution planning and policy seams

* refactor server execution services and passthrough semantics

* fix batch and passthrough review findings

* fix: removed unused helper

* add release e2e curl scenarios

* improve audit coverage for passthrough and failures

* docs: snapshot current architecture

* fix file pagination and batch preparation regressions

* remove dead file pagination helpers

* tighten lint config and remove dead planning paths

* fix dashboard contribution calendar month alignment