Skip to content

Releases: GoogleCloudPlatform/kubectl-ai

v0.0.12

04 Jun 21:58
d3d6951
Compare
Choose a tag to compare

What's Changed

  • Introduce MCP Client with Server Management and Tool Exec by @tuannvm in #270
  • Fix error when trying to include multi-line string in conversation output when using --enable-tool-use-shim by @noahlwest in #297
  • Add HTTP and stdio MCP Client Implementations with Unified Interface by @tuannvm in #291
  • fix and improve the observation format by @denverdino in #308
  • feat: Update google.golang.org/genai to v1.8.0 by @droot in #301
  • Fix NPE in conversation.go by @denverdino in #312
  • docs(README): add note on MCP Client Mode availability in v0.0.12 by @tuannvm in #313

New Contributors

Full Changelog: v0.0.11...v0.0.12

v0.0.11

28 May 21:25
0442b4a
Compare
Choose a tag to compare

What's Changed

  • gollm: bump default model for ollama provider by @ameukam in #261
  • Update OpenAI and Stream Handling for Compatibility with v1.0.0 API by @tuannvm in #262
  • eval: multi container pod communication with sidecar by @snewcomer in #253
  • FIX | Stop Agentic Loop and Await User Input When Tool Call is Declined by @zvdy in #268
  • cleanup: deduplicate common instruction logic in systemprompt_template_default.txt by @yankay in #272
  • Fixed tool config handling by @droot in #271
  • fix: improve stdin data detection logic in debug mode by @mikebz in #277
  • fix: Improve CLI prompt and input handling for confirmation choices by @THE-BRAHMA in #278
  • Created a Taskfile to help with routine tasks. by @mikebz in #279
  • Fix: make config.yaml Optional by @yankay in #281
  • feat(cli): Adding a Description to UserInterface in the CLI by @yankay in #283
  • FIX: kubectl streaming and interaction bashtool #122 #232 by @zvdy in #273
  • Revert "fix: improve stdin data detection logic in debug mode" by @droot in #284
  • feat: add UIListenAddress option for HTML UI configuration by @prasad89 in #287

New Contributors

Full Changelog: v0.0.10...v0.0.11

v0.0.10

20 May 22:10
dd52575
Compare
Choose a tag to compare

What's Changed

  • Document custom tools and enhance README by @janetkuo in #236
  • refactor(k8s-bench/tasks): update shebang in k8s-bench task scripts by @ruinshe in #237
  • fix(ci): Ensure job runs only for the correct repository by @mikebz in #241
  • feat: refactor OpenAI client to support customizable connection options by @appleboy in #243
  • Add modifies_resource parameter to custom tool by @janetkuo in #242
  • ci: add workflow for k8s-bench evaluations on push/PR by @tuannvm in #229
  • Add special command for listing available tools by @hakman in #247
  • Pass command args to custom tools by @hakman in #248
  • Support multiple custom tools configs by @hakman in #246
  • Optimize listing models for Azure OpenAI by @hakman in #255
  • Feat/#252 version handler by @zvdy in #256

Full Changelog: v0.0.9...v0.0.10

v0.0.9

16 May 00:34
6259a3d
Compare
Choose a tag to compare

What's Changed

  • Add support for listing models for Azure OpenAI by @hakman in #165
  • azopenai: fix the response content validation error by @yumosx in #169
  • modelserving: add distributed serving by @justinsb in #75
  • feat(install): add quick and manual install options in README and script by @tuannvm in #150
  • Docs(readme): fix the --quite argument by @gaga5lala in #172
  • chore: fix typos in kubectl-expect/main.go by @sadikkuzu in #176
  • k8s-bench | health probe troubleshooting eval task #145 by @zvdy in #163
  • Improve docs for contribution by @janetkuo in #183
  • chore(deps): update dependencies by @droot in #182
  • feat(k8s-bench): add scripts for HPA, rolling update, StatefulSet tasks by @tuannvm in #170
  • feat: implement ListModels function to retrieve model IDs from OpenAI by @Azathoth-X in #154
  • Chore: add existence check for build-images script by @seunggihong in #192
  • Doc enhancement: Add OLLAMA_HOST into example by @j796160836 in #194
  • fix Observable#AddSubscription on s is shadowed by @anyoptional in #186
  • gollm: add some logs for RegisterProvider by @yumosx in #191
  • doc: add krew install document by @LinPr in #185
  • Adjust summary of kubectl-ai capabilities by @janetkuo in #197
  • refactor(tools): use typed context keys for kubeconfig and work_dir by @Vinay-Khanagavi in #178
  • introduce cmd directory for the CLI code by @droot in #198
  • k8s-bench concurrent task execution by @zvdy in #199
  • User guide for the mcp-server mode by @janetkuo in #200
  • docs: Added pointer to mcp docs by @droot in #201
  • doc: fix typo in MCP doc by @titaneric in #219
  • Updated links in the install.sh script by @pilz0 in #212
  • Remove viper integration by @droot in #214
  • Add SHA-256 checksum verification to model download script by @seunggihong in #213
  • refactor: simplify AgentTextBlock methods and improve message handling by @mikebz in #195
  • Feature: add shields in README by @zvdy in #221
  • feat: add configurable SSL verification and refactor client options by @appleboy in #217
  • Add instructions to serve models with llama.cpp by @janetkuo in #202
  • doc: add VS Code MCP configuration by @titaneric in #224
  • Add support for additional prompt template files by @RRethy in #205
  • Enhance OpenAI Client Initialization and Model Handling with Environment Variable Support and Default Model Logic by @tuannvm in #207
  • add file extensions for discovery within IDEs and find. by @mikebz in #203
  • Lazy initialization of input readers and evals by @droot in #233
  • fix(tools): Use exec.LookPath to find bash executable before running bash commands by @ruinshe in #230
  • Run lightweight evals periodically by @droot in #234
  • Fixed cluster name in the eval job by @droot in #235
  • Add native streaming support for OpenAI clients by @zvdy in #215
  • feat: refactor provider registry for safety and improved diagnostics by @appleboy in #226
  • Support custom tools by @janetkuo in #223

New Contributors

Full Changelog: v0.0.8...v0.0.9

v0.0.8

07 May 18:43
a885d65
Compare
Choose a tag to compare

What's Changed

New Contributors

Big thanks to @mattn @zvdy @sinwoobang @justinsb @LinPr @selimacerbas for their contributions to this release. You'll are so awesome!

Full Changelog: v0.0.7...v0.0.8

v0.0.7

05 May 16:39
1afcbc4
Compare
Choose a tag to compare

Changelog

  • 38d629f Add usage info for Azure OpenAI (#135)
  • d2b52b0 Added readline support (#134)
  • b9148c0 Fix a mistake in README (#136)
  • 5fd397b Merge pull request #132 from mattn/fix-tmp2
  • f14757d Switched to cobra CLI framework and streamlined command line flags parsing
  • 1afcbc4 fixed the use of emojis in systemprompt_template (#138)
  • c64f297 use os.TempDir()

@selimacerbas @rxinui @cwrau @hakman @mattn @DoctorLai @mschneider82 contributed to this release. Thank you all 🎉

v0.0.6

30 Apr 17:01
ba8efb3
Compare
Choose a tag to compare

Changelog


Released by GoReleaser.

v0.0.5

28 Apr 21:17
feffcf4
Compare
Choose a tag to compare

Changelog


Released by GoReleaser.

v0.0.4

23 Apr 22:03
aeb3798
Compare
Choose a tag to compare

Changelog

  • aeb3798 Cleanup eval (#112)
  • 57dc924 Fix wrong indirect k8s.io/klog/v2 in the go.mod
  • a651944 Merge pull request #110 from zchee/k8s-bench_gomod-fmt

Released by GoReleaser.

v0.0.3

22 Apr 21:11
24eba2d
Compare
Choose a tag to compare

Changelog

  • 575d9a9 Add support for Azure OpenAI (#103)
  • 7ae286b Added demo gif (#102)
  • 6f1e425 Allow to specify the LLM provider ID instead of URL (#107)
  • 9e5da71 Fix the error building modelserving images
  • 24eba2d Fixed bug in create-network-policy eval (#109)
  • 8907a3a Pass the LLM provider URL to the Azure OpenAI client (#108)
  • 45c7e23 README for modelserving
  • b5f05da gollm: make it easier to build vertex client (#105)

Released by GoReleaser.