deepagents-code==0.1.67

### Features

- Added `GPT-6 Astra` support ([#6141](https://github.com/langchain-ai/deepagents/issues/6141)).
- Reuses the server across workspace switches ([#6152](https://github.com/langchain-ai/deepagents/issues/6152)).
- Limits thread resume age ([#6068](https://github.com/langchain-ai/deepagents/issues/6068)).

### Bug Fixes

- Improved workspace isolation by resolving project policy per workspace, scoping project dotenv to workspaces, and isolating LangSmith auth and workspace tracing ([#6064](https://github.com/langchain-ai/deepagents/issues/6064), [#5980](https://github.com/langchain-ai/deepagents/issues/5980), [#6061](https://github.com/langchain-ai/deepagents/issues/6061)).
- Improved transcript tail reconciliation by bounding and serializing reconciliation work ([#6057](https://github.com/langchain-ai/deepagents/issues/6057), [#6143](https://github.com/langchain-ai/deepagents/issues/6143)).
- Refuses a second sandbox workspace and moves sandbox creation and entry off the guarded event loop ([#5979](https://github.com/langchain-ai/deepagents/issues/5979), [#6017](https://github.com/langchain-ai/deepagents/issues/6017)).
- Surfaces stale updates in-session ([#6063](https://github.com/langchain-ai/deepagents/issues/6063)).
- Surfaces the LangSmith credential in `dcode config` ([#6083](https://github.com/langchain-ai/deepagents/issues/6083)).
- Offloads server config imports ([#6089](https://github.com/langchain-ai/deepagents/issues/6089)).
- Correctly parses extra skill directories with the platform separator ([#6151](https://github.com/langchain-ai/deepagents/issues/6151)).

---

Thanks to our community contributors: @jack-michaud, @UniversePeak

**Special thanks** to everyone who reported the issues addressed in this release: @jack-michaud ([#6018](https://github.com/langchain-ai/deepagents/issues/6018)), @JessYanCoding ([#6074](https://github.com/langchain-ai/deepagents/issues/6074))

Internal maintainers: @mdrxy

Released by: @mdrxy

<details>
<summary>Git log since deepagents-code==0.1.66</summary>

This commit history includes changes to this package. Commits are listed newest first.

- [`d0fbc7c`](https://github.com/langchain-ai/deepagents/commit/d0fbc7c16c233bc8e65e56e4be485d72c7a8abb2) release(deepagents-code): 0.1.67 (#6059)
- [`c71cd51`](https://github.com/langchain-ai/deepagents/commit/c71cd511841358c318b8fa6aa3222ecfd7890a64) chore(deps): bump httpx2 from 2.10.0 to 2.12.0 in /libs/code (#6179)
- [`f6ae37c`](https://github.com/langchain-ai/deepagents/commit/f6ae37cb1f20486f2da9d5a304681ea96de99bc1) chore(deps): raise dependency minimums for `all` (#6186)
- [`32fe5e4`](https://github.com/langchain-ai/deepagents/commit/32fe5e4b42a91bd2c8e2576510aa1dc55196c57d) chore(deps): bump httpcore2 from 2.9.0 to 2.10.0 in /libs/code (#6182)
- [`fd28bfb`](https://github.com/langchain-ai/deepagents/commit/fd28bfb5c5e1ca18af975d8e57d361c1c23f81b7) fix(code): run sandbox creation and entry off the guarded event loop (#6017)
- [`318311f`](https://github.com/langchain-ai/deepagents/commit/318311f56c3d3b9c4af4ece22fcfc387eeece5cc) feat(code): limit thread resume age (#6068)
- [`56e4ef1`](https://github.com/langchain-ai/deepagents/commit/56e4ef1a5f447950a04ba37a7855c8e0babd2f11) fix(code): parse extra skill dirs with platform separator (#6151)
- [`ca2d1ef`](https://github.com/langchain-ai/deepagents/commit/ca2d1ef7f270d0fa1040e5473f43e99937b5f03f) feat(code): reuse server across workspace switches (#6152)
- [`5fb8b69`](https://github.com/langchain-ai/deepagents/commit/5fb8b699843ecf93debc592d9dbb8a534f54db09) fix(code): serialize transcript tail reconciliation (#6143)
- [`a1a0887`](https://github.com/langchain-ai/deepagents/commit/a1a0887f7e42d0f8515ad415c9b9062e89350e8c) fix(code): surface the LangSmith credential in `dcode config` (#6083)
- [`ea1f253`](https://github.com/langchain-ai/deepagents/commit/ea1f2536537efb01c71dce54c46894b9318c4f80) fix(code): resolve project policy per workspace (#6064)
- [`794eda8`](https://github.com/langchain-ai/deepagents/commit/794eda83b777a2a7278e5726607e4b305801298d) fix(code): isolate LangSmith auth and workspace tracing (#6061)
- [`048f989`](https://github.com/langchain-ai/deepagents/commit/048f9890bfe8b640e1e72085a4fa50d9084545f0) feat(code): add `GPT-6 Astra` support (#6141)
- [`90fe4f4`](https://github.com/langchain-ai/deepagents/commit/90fe4f4ac0a1311f1527b1a5cd02bd3ac8664580) fix(code): offload server config import (#6089)
- [`18106be`](https://github.com/langchain-ai/deepagents/commit/18106be837bcdd9005b2dd73280d871a0621bf99) chore(deps): raise dependency minimums for `all` (#6130)
- [`07d2952`](https://github.com/langchain-ai/deepagents/commit/07d2952d346d81d06bd181db8c560a77f2b51bc8) chore(deps): raise dependency minimums for `all` (#6096)
- [`6c89fe2`](https://github.com/langchain-ai/deepagents/commit/6c89fe2197a2dfe4f3851cda38565bcadba6066b) release(langchain-quickjs): 0.3.7 (#6095)
- [`cdb9d4c`](https://github.com/langchain-ai/deepagents/commit/cdb9d4c30f4419c0eb2c98ad53940875e4dfbe9e) fix(code): surface stale updates in-session (#6063)
- [`678c115`](https://github.com/langchain-ai/deepagents/commit/678c11559d7534ee5e16ceebfacf01eeb6536ac7) fix(code): scope project dotenv to workspaces (#5980)
- [`cb4c973`](https://github.com/langchain-ai/deepagents/commit/cb4c973172f7838eefa328ea0c6ca334b6aaab98) fix(code): refuse a second sandbox workspace (#5979)
- [`8da3264`](https://github.com/langchain-ai/deepagents/commit/8da32644dc7dcad1125a840bc17ec3d5ea66b8c3) fix(code): bound transcript tail reconciliation (#6057)

</details>