chore: release main (headroomlabs-ai#931)

🤖 I have created a release *beep* *boop*
---


<details><summary>0.26.0</summary>

##
[0.26.0](headroomlabs-ai/headroom@v0.25.0...v0.26.0)
(2026-06-16)


### Features

* add Copilot BYOK provider wrapper utilities and CLI support
([headroomlabs-ai#1041](headroomlabs-ai#1041))
([1d0335f](headroomlabs-ai@1d0335f))
* add dashboard agent usage stats
([headroomlabs-ai#814](headroomlabs-ai#814))
([d8967bd](headroomlabs-ai@d8967bd))
* Add support for Mistral Vibe CLI
([headroomlabs-ai#935](headroomlabs-ai#935))
([38b5af6](headroomlabs-ai@38b5af6))
* attribute reread waste to over-compression via marker check
([headroomlabs-ai#901](headroomlabs-ai#901))
([77c0ec8](headroomlabs-ai@77c0ec8))
* **bedrock:** cross-region + Converse compression; bundle proxy binary
in images ([headroomlabs-ai#999](headroomlabs-ai#999))
([d036c4a](headroomlabs-ai@d036c4a))
* **dashboard:** surface compression-vs-cache net impact in Prefix Cache
panel ([headroomlabs-ai#913](headroomlabs-ai#913))
([b422a9e](headroomlabs-ai@b422a9e))
* **evals:** adversarial-input robustness grid for compressors
([headroomlabs-ai#918](headroomlabs-ai#918))
([7c51254](headroomlabs-ai@7c51254))
* **parser:** detect re-issued identical tool calls as reread waste
([headroomlabs-ai#909](headroomlabs-ai#909))
([d828b57](headroomlabs-ai@d828b57))
* **policy:** batch deep edits through one cache-bust
([headroomlabs-ai#856](headroomlabs-ai#856) P3a)
([headroomlabs-ai#1015](headroomlabs-ai#1015))
([2168943](headroomlabs-ai@2168943))
* **policy:** consume net-cost mutation gate in ContentRouter
([headroomlabs-ai#856](headroomlabs-ai#856) P2)
([headroomlabs-ai#905](headroomlabs-ai#905))
([38b3cb4](headroomlabs-ai@38b3cb4))
* **proxy:** compress AWS Bedrock InvokeModel requests via configurable
upstream ([headroomlabs-ai#720](headroomlabs-ai#720))
([c105902](headroomlabs-ai@c105902))


### Bug Fixes

* **anthropic:** strip styled Claude model ids
([headroomlabs-ai#651](headroomlabs-ai#651))
([b7bbdc9](headroomlabs-ai@b7bbdc9))
* **anyllm:** forward openai api_base/api_key to the any-llm backend
([headroomlabs-ai#942](headroomlabs-ai#942))
([headroomlabs-ai#954](headroomlabs-ai#954))
([343f47c](headroomlabs-ai@343f47c))
* **cache:** guard None exemplar embeddings in dynamic detector
([headroomlabs-ai#950](headroomlabs-ai#950))
([82ec599](headroomlabs-ai@82ec599))
* **cache:** name the missing piece in semantic detector guard
([headroomlabs-ai#1018](headroomlabs-ai#1018))
([dce6749](headroomlabs-ai@dce6749))
* **ci:** check out repo in PR Governance label job
([headroomlabs-ai#1021](headroomlabs-ai#1021))
([2fc7e08](headroomlabs-ai@2fc7e08))
* **ci:** make PR governance advisory
([headroomlabs-ai#1047](headroomlabs-ai#1047))
([629fef0](headroomlabs-ai@629fef0))
* **codex:** compute waste signals on the OpenAI Responses path
([headroomlabs-ai#898](headroomlabs-ai#898))
([b9f7c1e](headroomlabs-ai@b9f7c1e))
* **codex:** poll /wham/usage for subscription limits (handshake no
longer sends x-codex-* headers)
([headroomlabs-ai#924](headroomlabs-ai#924))
([852598e](headroomlabs-ai@852598e))
* **codex:** PR health label check state
([headroomlabs-ai#986](headroomlabs-ai#986))
([b231076](headroomlabs-ai@b231076))
* **codex:** retag thread providers so history menu stays whole across
the proxy boundary
([headroomlabs-ai#1034](headroomlabs-ai#1034))
([7f045a6](headroomlabs-ai@7f045a6))
* **codex:** write canonical hooks feature flag and migrate deprecated
codex_hooks ([headroomlabs-ai#743](headroomlabs-ai#743))
([c85a7c3](headroomlabs-ai@c85a7c3))
* **compression:** convert tree-sitter byte offsets to char offsets
([headroomlabs-ai#892](headroomlabs-ai#892))
([7a1bb9c](headroomlabs-ai@7a1bb9c))
* **compression:** correct JSON array item counting and entropy gate
([headroomlabs-ai#887](headroomlabs-ai#887))
([0442b2e](headroomlabs-ai@0442b2e))
* **compression:** keep container bodies compressible in code handler
([headroomlabs-ai#890](headroomlabs-ai#890))
([54d19cc](headroomlabs-ai@54d19cc))
* **compression:** measure short-value threshold on payload, not token
([headroomlabs-ai#889](headroomlabs-ai#889))
([4732d5e](headroomlabs-ai@4732d5e))
* **compression:** use thread-local tree-sitter parsers in code handler
([headroomlabs-ai#893](headroomlabs-ai#893))
([56fe715](headroomlabs-ai@56fe715))
* **gemini:** surface functionResponse payloads to waste-signal
detection ([headroomlabs-ai#897](headroomlabs-ai#897))
([8bf9bbb](headroomlabs-ai@8bf9bbb))
* **learn:** decode directory names with spaces in Windows project paths
([headroomlabs-ai#997](headroomlabs-ai#997))
([headroomlabs-ai#1027](headroomlabs-ai#1027))
([f227015](headroomlabs-ai@f227015))
* **learn:** scan subagent and workflow transcripts
([headroomlabs-ai#1045](headroomlabs-ai#1045))
([d39ec02](headroomlabs-ai@d39ec02))
* **openclaw:** declare headroom_retrieve tool contract
([headroomlabs-ai#947](headroomlabs-ai#947))
([f2dd153](headroomlabs-ai@f2dd153))
* **policy:** correct warm-cache penalty in net_mutation_gain to (S +
dT) ([headroomlabs-ai#903](headroomlabs-ai#903))
([b4b1f90](headroomlabs-ai@b4b1f90))
* **proxy:** add native Bedrock converse-stream route
([headroomlabs-ai#917](headroomlabs-ai#917))
([0839329](headroomlabs-ai@0839329))
* **proxy:** keep codex image-generation WS turns alive through the
relay ([headroomlabs-ai#1000](headroomlabs-ai#1000))
([8ee761d](headroomlabs-ai@8ee761d))
* **proxy:** make budget enforcement actually work
([headroomlabs-ai#885](headroomlabs-ai#885))
([1f5489b](headroomlabs-ai@1f5489b))
* **proxy:** read RTK gain stats globally by default
([headroomlabs-ai#957](headroomlabs-ai#957))
([7aab9a6](headroomlabs-ai@7aab9a6))
* route v1internal code assist requests to cloudcode-pa.googleapis…
([headroomlabs-ai#821](headroomlabs-ai#821))
([2a6d1ae](headroomlabs-ai@2a6d1ae))
* **serena:** stop the Serena dashboard popup and make --no-serena
actually disable Serena
([headroomlabs-ai#1003](headroomlabs-ai#1003))
([a0f9da3](headroomlabs-ai@a0f9da3))
* support Copilot Business subscription auth
([headroomlabs-ai#641](headroomlabs-ai#641))
([8995fcb](headroomlabs-ai@8995fcb))
* wire HEADROOM_EXCLUDE_TOOLS / HEADROOM_TOOL_PROFILES into Click proxy
entrypoint ([headroomlabs-ai#943](headroomlabs-ai#943))
([cc913e8](headroomlabs-ai@cc913e8))
* **wrap:** avoid duplicate top-level keys when injecting codex provider
([headroomlabs-ai#884](headroomlabs-ai#884))
([5c686e2](headroomlabs-ai@5c686e2))


### Code Refactoring

* DRY cache logic, add thread safety, fix Bash exclusion
([headroomlabs-ai#704](headroomlabs-ai#704))
([19ef66b](headroomlabs-ai@19ef66b))
</details>

---
This PR was generated with [Release
Please](https://github.com/googleapis/release-please). See
[documentation](https://github.com/googleapis/release-please#release-please).

Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>

Author: github-actions[bot]

.release-please-manifest.json

@@ -1,3 +1,3 @@
 {
-  ".": "0.25.0"
+  ".": "0.26.0"
 }

CHANGELOG.md

@@ -32,6 +32,62 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 * **wrap (codex):** fix `headroom wrap codex` producing a `config.toml` with duplicate top-level `model_provider` / `openai_base_url` keys (TOML-spec error) when the user had already configured their own provider. The injector now rewrites pre-existing top-level `model_provider` and `openai_base_url` lines in place — the previous value is kept in a `# was: …` trailing comment — instead of unconditionally prepending a duplicate, so `codex` can start against the proxy. The pre-wrap snapshot mechanism continues to byte-for-byte restore the original file on `headroom unwrap codex`.
 
 
+## [0.26.0](https://github.com/chopratejas/headroom/compare/v0.25.0...v0.26.0) (2026-06-16)
+
+
+### Features
+
+* add Copilot BYOK provider wrapper utilities and CLI support ([#1041](https://github.com/chopratejas/headroom/issues/1041)) ([e67ee2a](https://github.com/chopratejas/headroom/commit/e67ee2af658bce35fb4c71b45a0c5b294d7dcfdc))
+* add dashboard agent usage stats ([#814](https://github.com/chopratejas/headroom/issues/814)) ([6d3f39f](https://github.com/chopratejas/headroom/commit/6d3f39f213f4eb2d1c6c814b34e1bf6fe2a5c959))
+* Add support for Mistral Vibe CLI ([#935](https://github.com/chopratejas/headroom/issues/935)) ([0932b8b](https://github.com/chopratejas/headroom/commit/0932b8bef4db9109665382b6d7c079a368f08d52))
+* attribute reread waste to over-compression via marker check ([#901](https://github.com/chopratejas/headroom/issues/901)) ([f928576](https://github.com/chopratejas/headroom/commit/f9285766dda77b116c7834165849264e55339720))
+* **bedrock:** cross-region + Converse compression; bundle proxy binary in images ([#999](https://github.com/chopratejas/headroom/issues/999)) ([0dc2e1c](https://github.com/chopratejas/headroom/commit/0dc2e1cb3f7278332d450644831007316d6ac18c))
+* **dashboard:** surface compression-vs-cache net impact in Prefix Cache panel ([#913](https://github.com/chopratejas/headroom/issues/913)) ([2a4d300](https://github.com/chopratejas/headroom/commit/2a4d300841c8cbb55435f821fc2d01c3b3b43a59))
+* **evals:** adversarial-input robustness grid for compressors ([#918](https://github.com/chopratejas/headroom/issues/918)) ([5939004](https://github.com/chopratejas/headroom/commit/5939004185a1f9b4ef2e88ee3e72a10e5c8fa4a6))
+* **parser:** detect re-issued identical tool calls as reread waste ([#909](https://github.com/chopratejas/headroom/issues/909)) ([7d4ae86](https://github.com/chopratejas/headroom/commit/7d4ae86ec0bb09efff765422b89db587b050cd08))
+* **policy:** batch deep edits through one cache-bust ([#856](https://github.com/chopratejas/headroom/issues/856) P3a) ([#1015](https://github.com/chopratejas/headroom/issues/1015)) ([c2e52fe](https://github.com/chopratejas/headroom/commit/c2e52fe7439b464edaee83827ca7d8c8091d7e9a))
+* **policy:** consume net-cost mutation gate in ContentRouter ([#856](https://github.com/chopratejas/headroom/issues/856) P2) ([#905](https://github.com/chopratejas/headroom/issues/905)) ([553ade4](https://github.com/chopratejas/headroom/commit/553ade4ec66793c1707df6a95888ca2c1506c0b1))
+* **proxy:** compress AWS Bedrock InvokeModel requests via configurable upstream ([#720](https://github.com/chopratejas/headroom/issues/720)) ([7edb27a](https://github.com/chopratejas/headroom/commit/7edb27ab2496b070cbe835b31eb2f828798ddfaa))
+
+
+### Bug Fixes
+
+* **anthropic:** strip styled Claude model ids ([#651](https://github.com/chopratejas/headroom/issues/651)) ([0c5c89d](https://github.com/chopratejas/headroom/commit/0c5c89d05cefabaa833e54decfdeb677edacc0d7))
+* **anyllm:** forward openai api_base/api_key to the any-llm backend ([#942](https://github.com/chopratejas/headroom/issues/942)) ([#954](https://github.com/chopratejas/headroom/issues/954)) ([a7ee8a6](https://github.com/chopratejas/headroom/commit/a7ee8a60a7ac28a8adcc7a7fa83a04a59afe41d5))
+* **cache:** guard None exemplar embeddings in dynamic detector ([#950](https://github.com/chopratejas/headroom/issues/950)) ([1ec9320](https://github.com/chopratejas/headroom/commit/1ec93208883f2606cc7ec3db0b8bd8e071646984))
+* **cache:** name the missing piece in semantic detector guard ([#1018](https://github.com/chopratejas/headroom/issues/1018)) ([3b0bcee](https://github.com/chopratejas/headroom/commit/3b0bceecf4281eb34112de8dd546d4a58beb3fcc))
+* **ci:** check out repo in PR Governance label job ([#1021](https://github.com/chopratejas/headroom/issues/1021)) ([4558bc2](https://github.com/chopratejas/headroom/commit/4558bc2465e52d575070e5a0d6312cd400c8aee1))
+* **ci:** make PR governance advisory ([#1047](https://github.com/chopratejas/headroom/issues/1047)) ([74dff94](https://github.com/chopratejas/headroom/commit/74dff94fb8580426f5713991be71df94c4f31598))
+* **codex:** compute waste signals on the OpenAI Responses path ([#898](https://github.com/chopratejas/headroom/issues/898)) ([b9e2761](https://github.com/chopratejas/headroom/commit/b9e27614c613a1e5f97eb51af74d3c796fb1ab18))
+* **codex:** poll /wham/usage for subscription limits (handshake no longer sends x-codex-* headers) ([#924](https://github.com/chopratejas/headroom/issues/924)) ([8c00f71](https://github.com/chopratejas/headroom/commit/8c00f7103cf0288991d703cc002ac354e6266534))
+* **codex:** PR health label check state ([#986](https://github.com/chopratejas/headroom/issues/986)) ([99c874d](https://github.com/chopratejas/headroom/commit/99c874d4233ec2d35c5c12a709ba32fd2fd96f3d))
+* **codex:** retag thread providers so history menu stays whole across the proxy boundary ([#1034](https://github.com/chopratejas/headroom/issues/1034)) ([74ae781](https://github.com/chopratejas/headroom/commit/74ae7816444ae972b55f3da0ff5e28c8638ab4f3))
+* **codex:** write canonical hooks feature flag and migrate deprecated codex_hooks ([#743](https://github.com/chopratejas/headroom/issues/743)) ([dff6a19](https://github.com/chopratejas/headroom/commit/dff6a19946b8f96bb8b16fa945b69a1ed09709af))
+* **compression:** convert tree-sitter byte offsets to char offsets ([#892](https://github.com/chopratejas/headroom/issues/892)) ([b1f700f](https://github.com/chopratejas/headroom/commit/b1f700fc275bf1d7e9461b61a9ebfdb1fba19620))
+* **compression:** correct JSON array item counting and entropy gate ([#887](https://github.com/chopratejas/headroom/issues/887)) ([d6f0f0f](https://github.com/chopratejas/headroom/commit/d6f0f0f64269bfbdf36070cb304703c606c64b72))
+* **compression:** keep container bodies compressible in code handler ([#890](https://github.com/chopratejas/headroom/issues/890)) ([16ed73b](https://github.com/chopratejas/headroom/commit/16ed73bca68e602a86a385480d484c3a60025b8c))
+* **compression:** measure short-value threshold on payload, not token ([#889](https://github.com/chopratejas/headroom/issues/889)) ([65b0e8c](https://github.com/chopratejas/headroom/commit/65b0e8c58dbbc0b77e4b7159b279287979767c4c))
+* **compression:** use thread-local tree-sitter parsers in code handler ([#893](https://github.com/chopratejas/headroom/issues/893)) ([6cdb846](https://github.com/chopratejas/headroom/commit/6cdb8462000d9610b5d15f6c7c45adb787bfec1e))
+* **gemini:** surface functionResponse payloads to waste-signal detection ([#897](https://github.com/chopratejas/headroom/issues/897)) ([9b0c840](https://github.com/chopratejas/headroom/commit/9b0c840dd7c181d6266b31cd16f493393ccc5c1a))
+* **learn:** decode directory names with spaces in Windows project paths ([#997](https://github.com/chopratejas/headroom/issues/997)) ([#1027](https://github.com/chopratejas/headroom/issues/1027)) ([2d3701b](https://github.com/chopratejas/headroom/commit/2d3701b59e9ff8aedc2a282c4467f27ca2355d62))
+* **learn:** scan subagent and workflow transcripts ([#1045](https://github.com/chopratejas/headroom/issues/1045)) ([0ddd4ed](https://github.com/chopratejas/headroom/commit/0ddd4ed9e92fe898373036ba3be228f9afc3bc5a))
+* **openclaw:** declare headroom_retrieve tool contract ([#947](https://github.com/chopratejas/headroom/issues/947)) ([7c8c909](https://github.com/chopratejas/headroom/commit/7c8c909c853a264c833c645403cbbb1894b91432))
+* **policy:** correct warm-cache penalty in net_mutation_gain to (S + dT) ([#903](https://github.com/chopratejas/headroom/issues/903)) ([0632eba](https://github.com/chopratejas/headroom/commit/0632eba6c3bdf5b030d794d3dfefa3c29543d2e8))
+* **proxy:** add native Bedrock converse-stream route ([#917](https://github.com/chopratejas/headroom/issues/917)) ([b08ec15](https://github.com/chopratejas/headroom/commit/b08ec15b0d392b8b8cf93dbadaee4b7e6b465f1c))
+* **proxy:** keep codex image-generation WS turns alive through the relay ([#1000](https://github.com/chopratejas/headroom/issues/1000)) ([7dbbb40](https://github.com/chopratejas/headroom/commit/7dbbb4077e7bb11b3da4634573cfc1d998e139ec))
+* **proxy:** make budget enforcement actually work ([#885](https://github.com/chopratejas/headroom/issues/885)) ([a14ab45](https://github.com/chopratejas/headroom/commit/a14ab45cf0e6e698c52a0efd0448ca7c8ba0b31f))
+* **proxy:** read RTK gain stats globally by default ([#957](https://github.com/chopratejas/headroom/issues/957)) ([b70fccb](https://github.com/chopratejas/headroom/commit/b70fccbe174e1adff0f52ceaf9bec0dcda0c73da))
+* route v1internal code assist requests to cloudcode-pa.googleapis… ([#821](https://github.com/chopratejas/headroom/issues/821)) ([e20f16b](https://github.com/chopratejas/headroom/commit/e20f16b1a65710f532aa019ef60ac7a18a4e7f46))
+* **serena:** stop the Serena dashboard popup and make --no-serena actually disable Serena ([#1003](https://github.com/chopratejas/headroom/issues/1003)) ([919379a](https://github.com/chopratejas/headroom/commit/919379a8a1731a0002d813a79d880ad35f8bbbc9))
+* support Copilot Business subscription auth ([#641](https://github.com/chopratejas/headroom/issues/641)) ([0b4a4bd](https://github.com/chopratejas/headroom/commit/0b4a4bd4830ecec1bca64c2f62455c4c923d91df))
+* wire HEADROOM_EXCLUDE_TOOLS / HEADROOM_TOOL_PROFILES into Click proxy entrypoint ([#943](https://github.com/chopratejas/headroom/issues/943)) ([9b7b436](https://github.com/chopratejas/headroom/commit/9b7b436b04118d6ec4dcaebafc1c82e03e786f27))
+* **wrap:** avoid duplicate top-level keys when injecting codex provider ([#884](https://github.com/chopratejas/headroom/issues/884)) ([dd22cfd](https://github.com/chopratejas/headroom/commit/dd22cfd72ad9265c25a95ef5536dc3d17e85dbbf))
+
+
+### Code Refactoring
+
+* DRY cache logic, add thread safety, fix Bash exclusion ([#704](https://github.com/chopratejas/headroom/issues/704)) ([e36fccd](https://github.com/chopratejas/headroom/commit/e36fccd8cfe6b963398d3d0fa1637a45bd6421af))
+
 ## [0.25.0](https://github.com/chopratejas/headroom/compare/v0.24.0...v0.25.0) (2026-06-12)
 
 

plugins/openclaw/package.json

@@ -1,6 +1,6 @@
 {
   "name": "headroom-openclaw",
-  "version": "0.25.0",
+  "version": "0.26.0",
   "description": "Headroom context compression plugin for OpenClaw — 70-90% token savings with zero LLM calls",
   "type": "module",
   "main": "./dist/index.js",

pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "maturin"
 
 [project]
 name = "headroom-ai"
-version = "0.25.0"
+version = "0.26.0"
 description = "The Context Optimization Layer for LLM Applications - Cut costs by 50-90%"
 readme = "README.md"
 license = "Apache-2.0"

sdk/typescript/package.json

@@ -1,6 +1,6 @@
 {
   "name": "headroom-ai",
-  "version": "0.25.0",
+  "version": "0.26.0",
   "description": "Compress LLM context. Save tokens. Fit more into every request.",
   "type": "module",
   "main": "./dist/index.cjs",