deepagents-cli==0.0.20

### Features

* Model switcher & arbitrary chat model support ([#1127](https://github.com/langchain-ai/deepagents/issues/1127)) ([28fc311](https://github.com/langchain-ai/deepagents/commit/28fc311da37881257e409149022f0717f78013ef))
* Non-interactive mode w/ shell allow-listing ([#909](https://github.com/langchain-ai/deepagents/issues/909)) ([433bd2c](https://github.com/langchain-ai/deepagents/commit/433bd2cb493d6c4b59f2833e4304eead0304195a))
  * `--quiet` flag to suppress non-agent output w/ `-n` ([#1201](https://github.com/langchain-ai/deepagents/issues/1201)) ([3e96792](https://github.com/langchain-ai/deepagents/commit/3e967926655cf5249a1bc5ca3edd48da9dd3061b))
* Built-in skills, ship `skill-creator` as first ([#1191](https://github.com/langchain-ai/deepagents/issues/1191)) ([42823a8](https://github.com/langchain-ai/deepagents/commit/42823a88d1eb7242a5d9b3eba981f24b3ea9e274))
  * Enrich built-in skill metadata with license and compatibility info ([#1193](https://github.com/langchain-ai/deepagents/issues/1193)) ([b8179c2](https://github.com/langchain-ai/deepagents/commit/b8179c23f9130c92cb1fb7c6b34d98cc32ec092a))
* Implement message queue ([#1197](https://github.com/langchain-ai/deepagents/issues/1197)) ([c4678d7](https://github.com/langchain-ai/deepagents/commit/c4678d7641785ac4f17045eb75d55f9dc44f37fe))
* Add docs link to `/help` ([#1098](https://github.com/langchain-ai/deepagents/issues/1098)) ([8f8fc98](https://github.com/langchain-ai/deepagents/commit/8f8fc98bd403d96d6ed95fce8906d9c881236613))
* Support custom working directories and LangSmith sandbox templates ([#1099](https://github.com/langchain-ai/deepagents/issues/1099)) ([21e7150](https://github.com/langchain-ai/deepagents/commit/21e715054ea5cf48cab05319b2116509fbacd899))


### Bug Fixes

* Terminal virtualize scrolling to stop perf issues ([#965](https://github.com/langchain-ai/deepagents/issues/965)) ([5633c82](https://github.com/langchain-ai/deepagents/commit/5633c825832a0e8bd645681db23e97af31879b65))
* Update splash thread ID on `/clear` ([#1204](https://github.com/langchain-ai/deepagents/issues/1204)) ([23651ed](https://github.com/langchain-ai/deepagents/commit/23651edbc236e4a68fb0d9496506e6293b836cd9))
* `-m` initial prompt submission ([#1184](https://github.com/langchain-ai/deepagents/issues/1184)) ([a702e82](https://github.com/langchain-ai/deepagents/commit/a702e82a0f61edbadd78eff6906ecde20b601798))
  * Align skill-creator example scripts with agent skills spec ([#1177](https://github.com/langchain-ai/deepagents/issues/1177)) ([199d176](https://github.com/langchain-ai/deepagents/commit/199d17676ac1bfee645908a6c58193291e522890))
  * Port skills behavior from SDK ([#1192](https://github.com/langchain-ai/deepagents/issues/1192)) ([ad9241d](https://github.com/langchain-ai/deepagents/commit/ad9241da6e7e23e4430756a1d5a3afb6c6bfebcc)), closes [#1189](https://github.com/langchain-ai/deepagents/issues/1189)
* Rewrite skills create template to match spec guidance ([#1178](https://github.com/langchain-ai/deepagents/issues/1178)) ([f08ad52](https://github.com/langchain-ai/deepagents/commit/f08ad520172bd114e4cebf69138a10cbf98e157a))
* Harden dictionary iteration and HITL fallback handling ([#1151](https://github.com/langchain-ai/deepagents/issues/1151)) ([8b21fc6](https://github.com/langchain-ai/deepagents/commit/8b21fc6105d808ad25c53de96f339ab21efb4474))
* Per-subcommand help screens, short flags, and skills enhancements ([#1190](https://github.com/langchain-ai/deepagents/issues/1190)) ([3da1e8b](https://github.com/langchain-ai/deepagents/commit/3da1e8bc20bf39aba80f6507b9abc2352de38484))
* Refactor summarization middleware ([#1138](https://github.com/langchain-ai/deepagents/issues/1138)) ([e87001e](https://github.com/langchain-ai/deepagents/commit/e87001eace2852c2df47095ffd2611f09fdda2f5))

---

Thanks to our contributors: @1Ninad ([LinkedIn](https://linkedin.com/in/ninad22/)), @sydney-runkle, @an-ny89, @saakshigupta2002, @eyurtsev, @mdrxy