Merge branch 'master' into add_promql_capability

.gitattributes

@@ -1 +1 @@
-holmes/.git_archival.json  export-subst
+holmes/.git_archival.json  export-subst

.github/workflows/build-binaries-and-brew.yaml

@@ -22,7 +22,7 @@ jobs:
       uses: actions/setup-python@v2
       with:
         python-version: '3.11'
-    
+
     - name: Install dependencies
       if: matrix.os != 'windows-latest'
       run: |
@@ -43,7 +43,7 @@ jobs:
       if: matrix.os == 'ubuntu-20.04'
       run: |
         sudo apt-get install -y binutils
-    
+
     - name: Update package version (Linux)
       if: matrix.os == 'ubuntu-20.04'
       run: sed -i 's/__version__ = .*/__version__ = "${{ github.ref_name }}"/g' holmes/__init__.py
@@ -67,7 +67,7 @@ jobs:
       # regarding the tiktoken part of the command, see https://github.com/openai/tiktoken/issues/80
       # regarding the litellm part of the command, see https://github.com/pyinstaller/pyinstaller/issues/8620#issuecomment-2186540504
       run: |
-        pyinstaller holmes.py --add-data 'holmes/plugins/runbooks/*:holmes/plugins/runbooks' --add-data 'holmes/plugins/prompts/*:holmes/plugins/prompts' --add-data 'holmes/plugins/toolsets/*:holmes/plugins/toolsets' --hidden-import=tiktoken_ext.openai_public --hidden-import=tiktoken_ext --hiddenimport litellm.llms.tokenizers --hiddenimport litellm.litellm_core_utils.tokenizers --collect-data litellm 
+        pyinstaller holmes.py --add-data 'holmes/plugins/runbooks/*:holmes/plugins/runbooks' --add-data 'holmes/plugins/prompts/*:holmes/plugins/prompts' --add-data 'holmes/plugins/toolsets/*:holmes/plugins/toolsets' --hidden-import=tiktoken_ext.openai_public --hidden-import=tiktoken_ext --hiddenimport litellm.llms.tokenizers --hiddenimport litellm.litellm_core_utils.tokenizers --collect-data litellm
         ls dist
 
     - name: Zip the application (Unix)
@@ -91,7 +91,7 @@ jobs:
       env:
         GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
       with:
-        upload_url: ${{ github.event.release.upload_url }} 
+        upload_url: ${{ github.event.release.upload_url }}
         asset_path: ./holmes-${{ matrix.os }}-${{ github.ref_name }}.zip
         asset_name: holmes-${{ matrix.os }}-${{ github.ref_name }}.zip
         asset_content_type: application/octet-stream
@@ -105,7 +105,7 @@ jobs:
   check-latest:
     needs: build
     runs-on: ubuntu-20.04
-    outputs: 
+    outputs:
       IS_LATEST: ${{ steps.check-latest.outputs.release == github.ref_name }}
     steps:
       - id: check-latest

.github/workflows/build-docker-images.yaml

@@ -77,7 +77,7 @@ jobs:
     # Note: this ignores the "Set as latest release" checkbox in the GitHub UI
     # it isn't possible to check whether that was set or not
     # so if you do not want to override the "latest" tag, you should mark the release as a prerelease or a draft
-    # for prereleases and drafts we don't tag latest 
+    # for prereleases and drafts we don't tag latest
     - name: Tag and push Docker image as latest if applicable
       if: ${{ github.event.release.prerelease == false && github.event.release.draft == false }}
       run: |

.pre-commit-config.yaml

@@ -6,4 +6,16 @@ repos:
       - id: poetry-lock
         pass_filenames: false
         args:
-          - --no-update
+          - --no-update
+  - repo: https://github.com/astral-sh/ruff-pre-commit
+    rev: v0.7.2
+    hooks:
+      - id: ruff
+        entry: ruff check --fix
+      - id: ruff-format
+  - repo: https://github.com/pre-commit/pre-commit-hooks
+    rev: v5.0.0
+    hooks:
+      - id: detect-private-key
+      - id: end-of-file-fixer
+      - id: trailing-whitespace

CONTRIBUTING.md

@@ -15,15 +15,15 @@ Please make sure to read and observe our [Code of Conduct](https://github.com/ro
 
 ## Reporting bugs
 
-We encourage those interested to contribute code and also appreciate when issues are reported. 
+We encourage those interested to contribute code and also appreciate when issues are reported.
 
 - Create a new issue and label is as `bug`
 - Clearly state how to reproduce the bug:
   - Which LLM you've used
   - Which steps are required to reproduce
     - As LLMs answers may differ between runs - Does it always reproduce, or occasionally?
 
-  
+
 ## Contributing Code
 
 - Fork the repository and clone it locally.

Dockerfile

@@ -26,22 +26,34 @@ ENV PATH="$VIRTUAL_ENV/bin:$PATH"
 RUN curl -fsSL https://pkgs.k8s.io/core:/stable:/v1.29/deb/Release.key -o Release.key
 
 # Set the architecture-specific kube lineage URLs
-ARG ARM_URL=https://github.com/Avi-Robusta/kube-lineage/releases/download/v2.2.1/kube-lineage-macos-latest-v2.2.1
-ARG AMD_URL=https://github.com/Avi-Robusta/kube-lineage/releases/download/v2.2.1/kube-lineage-ubuntu-latest-v2.2.1
+ARG KUBE_LINEAGE_ARM_URL=https://github.com/Avi-Robusta/kube-lineage/releases/download/v2.2.1/kube-lineage-macos-latest-v2.2.1
+ARG KUBE_LINEAGE_AMD_URL=https://github.com/Avi-Robusta/kube-lineage/releases/download/v2.2.1/kube-lineage-ubuntu-latest-v2.2.1
 # Define a build argument to identify the platform
 ARG TARGETPLATFORM
 # Conditional download based on the platform
 RUN if [ "$TARGETPLATFORM" = "linux/arm64" ]; then \
-    curl -L -o kube-lineage $ARM_URL; \
+    curl -L -o kube-lineage $KUBE_LINEAGE_ARM_URL; \
     elif [ "$TARGETPLATFORM" = "linux/amd64" ]; then \
-    curl -L -o kube-lineage $AMD_URL; \
+    curl -L -o kube-lineage $KUBE_LINEAGE_AMD_URL; \
     else \
     echo "Unsupported platform: $TARGETPLATFORM"; exit 1; \
     fi
 RUN chmod 777 kube-lineage
 RUN ./kube-lineage --version
 
-RUN curl -sSL -o argocd-linux-amd64 https://github.com/argoproj/argo-cd/releases/latest/download/argocd-linux-amd64
+# Set the architecture-specific argocd URLs
+ARG ARGOCD_ARM_URL=https://github.com/argoproj/argo-cd/releases/latest/download/argocd-linux-arm64
+ARG ARGOCD_AMD_URL=https://github.com/argoproj/argo-cd/releases/latest/download/argocd-linux-amd64
+# Conditional download based on the platform
+RUN if [ "$TARGETPLATFORM" = "linux/arm64" ]; then \
+    curl -L -o argocd $ARGOCD_ARM_URL; \
+    elif [ "$TARGETPLATFORM" = "linux/amd64" ]; then \
+    curl -L -o argocd $ARGOCD_AMD_URL; \
+    else \
+    echo "Unsupported platform: $TARGETPLATFORM"; exit 1; \
+    fi
+RUN chmod 777 argocd
+RUN ./argocd --help
 
 # Install Helm
 RUN curl https://baltocdn.com/helm/signing.asc | gpg --dearmor -o /usr/share/keyrings/helm.gpg \
@@ -100,8 +112,7 @@ COPY --from=builder /app/kube-lineage /usr/local/bin
 RUN kube-lineage --version
 
 # Set up ArgoCD
-COPY --from=builder /app/argocd-linux-amd64 /usr/local/bin/argocd
-RUN chmod 555 /usr/local/bin/argocd
+COPY --from=builder /app/argocd /usr/local/bin/argocd
 RUN argocd --help
 
 # Set up Helm

Dockerfile.dev

@@ -59,7 +59,7 @@ ARG PRIVATE_PACKAGE_REGISTRY="none"
 RUN if [ "${PRIVATE_PACKAGE_REGISTRY}" != "none" ]; then \
     pip config set global.index-url "${PRIVATE_PACKAGE_REGISTRY}"; \
     fi \
-    && pip install poetry    
+    && pip install poetry
 ARG POETRY_REQUESTS_TIMEOUT
 RUN poetry config virtualenvs.create false
 COPY pyproject.toml poetry.lock /app/

Makefile

@@ -0,0 +1,4 @@
+
+
+check:
+	poetry run pre-commit run -a

README.md

@@ -631,31 +631,31 @@ Using Grafana Loki
 
 HolmesGPT can consult logs from [Loki](https://grafana.com/oss/loki/) by proxying through a [Grafana](https://grafana.com/oss/grafana/) instance.
 
-There are 2 parts to configuring access to Grafana Loki: Access/Authentication and search terms.
+To configure loki toolset:
 
-For access and authentication, add the following environment variables:
-
-* `GRAFANA_URL` - e.g. https://my-org.grafana.net
-* `GRAFANA_API_KEY` - e.g. glsa_bsm6ZS_sdfs25f
+```yaml
+toolsets:
+  grafana/loki:
+    enabled: true
+    config:
+      api_key: "{{ env.GRAFANA_API_KEY }}"
+      url: "http://loki-url"
+```
 
 For search terms, you can optionally tweak the search terms used by the toolset.
-This is done by appending the following to your Holmes configuration file:
+This is done by appending the following to your Holmes grafana/loki configuration:
 
 ```yaml
-grafana:
-  url: https://my-org.grafana.net #
-  api_key: glsa_bsm6ZS_sdfs25f
-  loki:
-    pod_name_search_key: "pod"
-    namespace_search_key: "namespace"
-    node_name_search_key: "node"
+pod_name_search_key: "pod"
+namespace_search_key: "namespace"
+node_name_search_key: "node"
 ```
 
 > You only need to tweak the configuration file if your Loki logs settings for pod, namespace and node differ from the above defaults.
 
-The Loki toolset is configured the using the same Grafana settings as the Grafana Tempo toolset.
 </details>
 
+<details>
 <summary>
 Using Grafana Tempo
 </summary>
@@ -664,8 +664,6 @@ HolmesGPT can fetch trace information from Grafana Tempo to debug performance re
 
 Tempo is configured the using the same Grafana settings as the Grafana Loki toolset.
 
-grafana:
-  url: https://my-org.grafana.net #
 </details>
 
 
@@ -875,9 +873,9 @@ Configure Slack to send notifications to specific channels. Provide your Slack t
 <summary>OpenSearch Integration</summary>
 
 The OpenSearch toolset (`opensearch`) allows Holmes to consult an opensearch cluster for its health, settings and shards information.
-The toolset supports multiple opensearch or elasticsearch clusters that are configured by editing Holmes' configuration file (or in cluster to the configuration secret):
+The toolset supports multiple opensearch or elasticsearch clusters that are configured by editing Holmes' configuration file:
 
-```                                                                                 
+```
 opensearch_clusters:
   - hosts:
       - https://my_elasticsearch.us-central1.gcp.cloud.es.io:443

examples/custom_llm.py

@@ -1,17 +1,14 @@
-
 from typing import Any, Dict, List, Optional, Type, Union
-from holmes.config import Config
 from holmes.core.llm import LLM
 from litellm.types.utils import ModelResponse
 from holmes.core.tool_calling_llm import ToolCallingLLM
 from holmes.core.tools import Tool, ToolExecutor
 from holmes.plugins.toolsets import load_builtin_toolsets
-from rich.console import Console
 from pydantic import BaseModel
 from holmes.plugins.prompts import load_and_render_prompt
-import sys
-class MyCustomLLM(LLM):
 
+
+class MyCustomLLM(LLM):
     def get_context_window_size(self) -> int:
         return 128000
 
@@ -21,36 +18,41 @@ def get_maximum_output_token(self) -> int:
     def count_tokens_for_message(self, messages: list[dict]) -> int:
         return 1
 
-    def completion(self, messages: List[Dict[str, Any]], tools: Optional[List[Tool]] = [], tool_choice: Optional[Union[str, dict]] = None, response_format: Optional[Union[dict, Type[BaseModel]]] = None, temperature:Optional[float] = None, drop_params: Optional[bool] = None) -> ModelResponse:
-        return ModelResponse(choices=[{
-            "finish_reason": "stop",
-            "index": 0,
-            "message": {
-                "role": "assistant",
-                "content": "There are no issues with your cluster"
-            }
-        }],
-        usage={
-            "prompt_tokens": 0,       # Integer
-            "completion_tokens": 0,
-            "total_tokens": 0
-        }
-    )
+    def completion(
+        self,
+        messages: List[Dict[str, Any]],
+        tools: Optional[List[Tool]] = [],
+        tool_choice: Optional[Union[str, dict]] = None,
+        response_format: Optional[Union[dict, Type[BaseModel]]] = None,
+        temperature: Optional[float] = None,
+        drop_params: Optional[bool] = None,
+    ) -> ModelResponse:
+        return ModelResponse(
+            choices=[
+                {
+                    "finish_reason": "stop",
+                    "index": 0,
+                    "message": {
+                        "role": "assistant",
+                        "content": "There are no issues with your cluster",
+                    },
+                }
+            ],
+            usage={
+                "prompt_tokens": 0,  # Integer
+                "completion_tokens": 0,
+                "total_tokens": 0,
+            },
+        )
 
 
 def ask_holmes():
-    console = Console()
-
     prompt = "what issues do I have in my cluster"
 
     system_prompt = load_and_render_prompt("builtin://generic_ask.jinja2")
 
     tool_executor = ToolExecutor(load_builtin_toolsets())
-    ai = ToolCallingLLM(
-        tool_executor,
-        max_steps=10,
-        llm=MyCustomLLM()
-    )
+    ai = ToolCallingLLM(tool_executor, max_steps=10, llm=MyCustomLLM())
 
     response = ai.prompt_call(system_prompt, prompt)
 

examples/custom_runbooks.yaml

@@ -4,4 +4,4 @@ runbooks:
     instructions: >
       Analyze pod logs for errors and also read the monogodb logs
       Correlate between the two logs and try to find the root cause of the issue.
-      Based on the logs, report the session ids of impacted transactions
+      Based on the logs, report the session ids of impacted transactions

examples/custom_toolset.yaml

@@ -11,7 +11,7 @@ toolsets:
     docs_url: "https://kubernetes.io/docs/home/"
     # Icon URL. Used for display in the UI
     icon_url: "https://encrypted-tbn0.gstatic.com/images?q=tbn:ANd9GcRPKA-U9m5BxYQDF1O7atMfj9EMMXEoGu4t0Q&s"
-    # Tags for categorizing toolsets, 'core' will be used for all Holmes features (both cli's commands and chats in UI). 
+    # Tags for categorizing toolsets, 'core' will be used for all Holmes features (both cli's commands and chats in UI).
     # The 'cluster' tag is used for UI functionality, while 'cli' is for for command-line specific tools
     tags:
       - core
@@ -24,7 +24,7 @@ toolsets:
       - name: "switch_cluster"
         # The LLM looks at this description when deciding what tools are relevant for each task
         description: "Used to switch between multiple kubernetes contexts(clusters)"
-        
+
         # A templated bash command using Jinja2 templates
         # The LLM can only control parameters that you expose as template variables like {{ this_variable }}
         command: "kubectl config use-context {{ cluster_name }}"

helm/holmes/Chart.yaml

@@ -7,4 +7,4 @@ type: application
 # we use 0.0.1 as a placeholder for the version` because Helm wont allow `0.0.0` and we want to be able to run
 # `helm install` on development checkouts without updating this file. the version doesn't matter in that case anyway
 version: 0.0.1
-appVersion: 0.0.0
+appVersion: 0.0.0

helm/holmes/templates/holmesgpt-service-account.yaml

@@ -229,4 +229,4 @@ subjects:
   - kind: ServiceAccount
     name: {{ .Release.Name }}-holmes-service-account
     namespace: {{ .Release.Namespace }}
-{{- end }}
+{{- end }}

holmes/.git_archival.json

@@ -5,4 +5,3 @@
   "refs": "$Format:%D$",
   "describe": "$Format:%(describe:tags=true,match=v[0-9]*)$"
 }
-