0.7.2

What's Changed

• Llm eval improvements by @nherment in #214
• add option to fetch deployment config by @RoiGlinik in #222
• test: llm eval, add correctness score, port alertmanager test cases by @nherment in #216
• Use Contributor Covenant Code of Conduct v2.1 by @Sheeproid in #223
• add contributing doc page by @Sheeproid in #225
• Create MAINTAINERS.md by @Sheeproid in #226

Full Changelog: 0.7.1...0.7.2

0.7.1

What's Changed

• Add precommit by @moshemorad in #218

Full Changelog: 0.6.1...0.7.1

What's Changed

• Add tags to ask holmes by @nherment in #185
• Fix for deprecated /api/conversation endpoint by @itisallgood in #196
• Fix llm tests on github runners by @nherment in #198
• speed up re-building on code changes by changes caching order by @Sheeproid in #202
• Get top memory requests tool by @Sheeproid in #205
• add kubectl_logs_grep tool by @Sheeproid in #203
• MAIN-2381 add support for both variations. by @RoiGlinik in #201
• Add kubectl events capability & improve prompt by @nherment in #206
• Provide the LLM the ability to fetch all or specific container logs within a pod by @nherment in #207
• Improve logging by @aantn in #211
• Curl in base by @Sheeproid in #212
• Add llm evaluations by @nherment in #204
• Update README.md by @aantn in #213
• Make holmesgpt work with WatsonX by @itisallgood in #200
• holmes to use grep logs more by tweaking instruction, move to grep -E by @Sheeproid in #215
• Add precommit by @moshemorad in #218

Full Changelog: 0.6.1...0.7.1

0.6.1

What's Changed

• Fix build_chat_message when number of tools is 0 by @itisallgood in #186
• add holmes cluster role for reading storage classes by @arikalon1 in #187
• Avoid server crash if no instruction is passed to workload health request by @nherment in #188
• Remove broken link in readme by @aantn in #190
• fix access to document dict url attribute by @arikalon1 in #189
• Only load builtin toolsets when the config is initialised by @nherment in #193
• fix: incorrect prerequisite for webpage: use python -m to invoke playwright exec and specify browser by @nherment in #192
• use internet tool by default on the chart by @arikalon1 in #194

Full Changelog: 0.6.0...0.6.1

0.6.0

What's Changed

• MAIN-2339 remove old support check code and use drop params by @RoiGlinik in #158
• Broken link fix by @pavangudiwada in #159
• added command that is sometimes selected to get a single rds instance by @Sheeproid in #161
• Added kubectl top action by @Avi-Robusta in #160
• [MAIN-2367] moved it to its own toolset by @Avi-Robusta in #164
• MAIN-2364 by @itisallgood in #163
• feat: add support for passing runbook URLs to issue investigator by @nherment in #165
• Avoid listing tools used during investigations by @nherment in #166
• Update README.md by @aantn in #170
• Custom llm improvements by @Avi-Robusta in #169
• [MAIN-2451] updating starlette by @Avi-Robusta in #177
• bug fixes: by @arikalon1 in #178
• support removing bedrock prefix for model name token by @arikalon1 in #179
• Reduce image size by removing pandoc and using firefox instead of chromium by @nherment in #176
• Support custom LLM implementation by @nherment in #174
• fix bug with wrong import name by @arikalon1 in #181
• MAIN-2345 Support new conversation functionality by @itisallgood in #162
• Fixed import of PGAPIError by @itisallgood in #182
• Add resources requests & limit to charts by @moshemorad in #184

New Contributors

• @nherment made their first contribution in #165
• @moshemorad made their first contribution in #184

Full Changelog: 0.5.2...0.6.0

0.5.2

What's Changed

• Added fallback for token count by @itisallgood in #156

Full Changelog: 0.5.1...0.5.2

0.5.1

What's Changed

• Added backwards compatibility with model names that start with openai by @Avi-Robusta in #153

Full Changelog: 0.5.0...0.5.1

0.5.0

What's Changed

• Added custom toolsets to helm chart by @itisallgood in #147
• Added better auth in cluster by @arikalon1 in #149

Full Changelog: 0.4.0...0.5.0

0.4.0

What's Changed

• Fix wrong docker image in README.md by @aantn in #125
• Generated by running "poetry update litellm" by @aantn in #130
• Update README with Ollama instructions by @aantn in #133
• MAIN-2111 patching Holmes-CVE-2024-41110 by @Avi-Robusta in #132
• Now "/api/conversation" return tool_calls by @itisallgood in #134
• update litellm version to remove spammy error by @aantn in #135
• remove unused file by @aantn in #136
• remove unused part of ask template (server doesn't use this jinja2 te… by @aantn in #137
• Improvements to cli logging and --verbose flag by @aantn in #138
• Reuse prompts by @aantn in #139
• Handle large inputs and LLM context windows better for ask by @itisallgood in #131
• bug fix for verbosity by @aantn in #141
• MAIN-2244 Updated github workflows with new versions of download-artifact and upload-artifact by @itisallgood in #140
• MAIN-2239 fix expat CVE by @itisallgood in #142
• Update prompts to get more detailed output by @aantn in #143
• allow overriding the UI token with an env var. by @arikalon1 in #144
• MAIN-2220 remove old unnecessary code by @RoiGlinik in #146

Full Changelog: 0.3.0...0.4.0

0.3.0

What's Changed

• Fixing alert filter using alertname by @pavangudiwada in #93
• Corrected Pagerduty error message by @kmurthyms in #94
• Allow reading model param from config file by @aantn in #95
• Update litellm and other dependency versions by @aantn in #98
• Parallel tool calling by @aantn in #97
• Added docs for private servers by @Avi-Robusta in #99
• Increase accuracy of holmes ask results by @aantn in #89
• [MAIN-1849] added dockerhub build to gitaction by @Avi-Robusta in #96
• Fix typo in toolname in prompt by @aantn in #101
• add .idea to .gitignore by @Sheeproid in #108
• check prerequisites for kube-lineage & support built from source by @Sheeproid in #109
• update litellm to solve BerriAI/litellm#4933 by @Sheeproid in #110
• Added K9s integration docs by @pavangudiwada in #107
• comments and casing by @aantn in #103
• Fix "lazy" answers when investigating jira issues by @aantn in #105
• allow not echoing user request by @aantn in #102
• Minor improvements to K9s integration by @aantn in #111
• Updated K9s plugin by @pavangudiwada in #112
• CVE-2024-6345 fix by @itisallgood in #114
• Add AWS, confluence, internet and slab tools + minor improvements to general tool loading by @aantn in #104
• Trusting custom CA certificates by @arikalon1 in #115
• Update generate command for our internal testing purposes by @aantn in #117
• Main 1907 workload health endpoint by @RoiGlinik in #106
• Add the option to use robusta ai with holmes by @RoiGlinik in #116
• Holmes workload health with json response by @RoiGlinik in #122
• allow adding tolerations to the holmes pod by @arikalon1 in #124
• Chat input box api by @itisallgood in #118
• extended sent instructions and do not override them by @RoiGlinik in #126

New Contributors

• @kmurthyms made their first contribution in #94
• @Sheeproid made their first contribution in #108
• @RoiGlinik made their first contribution in #106

Full Changelog: 0.2.6...0.3.0

0.2.6

New Features

Simplified Integration with Amazon Bedrock and Other LLM Models

Integration with Amazon Bedrock and other large language models (LLMs) has been streamlined to provide a more seamless experience.
For more details, refer to PR #65.

Slack Destination for Ask Command

A new feature has been introduced allowing users to send queries directly to Slack using the ask command, enhancing communication and integration within Slack channels.
For more details, refer to PR #81.

Enhancements

• Update PagerDuty integration by @aantn in #86
• Minor improvements for testing by @aantn in #75
• Make it easier to test Alertmanager from our test framework by @aantn in #79
• Minor improvements to Kubernetes toolset by @aantn in #85
• Added Prometheus Operator RBAC for Holmes to service account template by @itisallgood in #87
• Update the Helm chart to support working with LiteLLM env vars by @arikalon1 in #92
• [MAIN-1860] Added a post-processing option for chatty AIs by @Avi-Robusta in #88

Security

• Security updates for Holmes by @Avi-Robusta in #83 and #90

New Contributors

• @itisallgood made their first contribution in #87

Full Changelog: 0.2.5...0.2.6