holmesgpt

Summary

HolmesGPT is an open-source AI agent, run from the command line, that investigates production incidents and finds root causes by querying live observability data across Kubernetes, VMs, cloud providers, databases, and SaaS platforms. It is aimed at SRE and platform teams handling alert triage, connecting bidirectionally to sources like Prometheus, Grafana, Datadog, AlertManager, PagerDuty, and Jira, and its new operator mode can run continuously in the background, flagging problems and messaging fixes to Slack before customers notice. It calls itself the CNCF SRE Agent, positioning it as a category of its own rather than a clone of an existing tool, built originally by Robusta.Dev with major contributions from Microsoft. With 90 contributors and over 3,000 commits in the past year, and a release just 15 days old, it is under active development.

SRE Agent - CNCF Sandbox Project

Command line

Values are placeholders taken from each option’s declared default. Nothing is executed here — the output shown is a recording of a run that already happened.

Release history

1. 0.40.0 Aug 26, 2026 · issue 008

   HolmesGPT 0.40.0 adds personal skills, per-account skill hierarchy, tcp_check allowlists, and diagnostic-pod target policies.

   └──▷ GET THIS VERSION
   copy

   $ git clone --branch 0.40.0 https://github.com/HolmesGPT/holmesgpt.git
   # already have the repo? check out this version:
   $ git checkout 0.40.0

   • ›Requires an allowlist before tcp_check probes can reach private networks, blocking SSRF to internal hosts by default.
   • ›Adds a diagnostic-pod target policy to prevent SSRF and data exfiltration from pods spawned during investigations.
   • ›Adds personal skills and a per-account name-collision hierarchy, so individual users can define skills that don't conflict with shared or global ones.
   • ›Validates bash toolset command arguments and redirections — not just the command prefix — closing injection vectors.
   • ›Hardens the kubectl-run toolset to execute commands without a host shell, eliminating shell injection paths.
   +8 moreshow less

   • ›Caps oversized error tool results: a 500-character gist is inlined and the full content is spilled to a file.
   • ›Surfaces skill titles for UI-added skills in fetch_skill results.
   • ›Injects __robusta_user_approved only into relay-routed remote MCP calls, not all MCP traffic.
   • ›Wires Kubernetes-remediation MCP bearer-token auth in the Helm chart.
   • ›Updates the GitLab MCP integration to v2.1.46 with a whoami health check.
   • ›Gracefully retires in-flight conversations on shutdown instead of dropping them.
   • ›Removes the experimental Azure SQL toolset.
   • ›Removes the MariaDB MCP integration, consolidating to the native MariaDB datasource.
   └──▷ BREAKING ON UPGRADE

   • !The MariaDB MCP integration has been removed; deployments using it must migrate to the native MariaDB datasource.
   • !The experimental Azure SQL toolset has been removed; any configuration referencing it will no longer load.
2. 0.39.0 Aug 10, 2026 · issue -009

   HolmesGPT 0.39.0 adds Atlassian Rovo MCP integration and GitHub App multi-org support for the GitHub MCP addon.

   └──▷ GET THIS VERSION
   copy

   $ git clone --branch 0.39.0 https://github.com/HolmesGPT/holmesgpt.git
   # already have the repo? check out this version:
   $ git checkout 0.39.0

   • ›Adds Atlassian Rovo MCP integration, enabling HolmesGPT to connect with Rovo as an MCP source.
   • ›Supports pulling skills using GitHub App credentials.
   • ›Supports GitHub App installations spanning multiple organizations in the GitHub MCP addon.
3. 0.39.0 Aug 10, 2026 · issue 001

   HolmesGPT 0.39.0 adds Atlassian Rovo MCP integration, GitHub App credential-based skill pulling, and multi-org GitHub App support.

   └──▷ GET THIS VERSION
   copy

   $ git clone --branch 0.39.0 https://github.com/HolmesGPT/holmesgpt.git
   # already have the repo? check out this version:
   $ git checkout 0.39.0

   • ›Extends GitHub MCP addon to support GitHub App installations spanning multiple organizations.
   • ›Adds Atlassian Rovo MCP integration, enabling access to Jira issues and Confluence pages via Atlassian's hosted MCP server.
4. 0.38.2 Aug 6, 2026 · issue -013

   HolmesGPT 0.38.2 adds RFC 8707 resource indicator support to the MCP OAuth flow.

   └──▷ GET THIS VERSION
   copy

   $ git clone --branch 0.38.2 https://github.com/HolmesGPT/holmesgpt.git
   # already have the repo? check out this version:
   $ git checkout 0.38.2

   • ›Adds RFC 8707 resource indicator support to the MCP OAuth flow, strengthening token audience binding during authorization.
5. 0.38.2 Aug 6, 2026 · issue 001

   HolmesGPT 0.38.2 adds RFC 8707 resource indicator support to the MCP OAuth flow.

   └──▷ GET THIS VERSION
   copy

   $ git clone --branch 0.38.2 https://github.com/HolmesGPT/holmesgpt.git
   # already have the repo? check out this version:
   $ git checkout 0.38.2

   • ›Adds RFC 8707 resource indicator support to the MCP OAuth flow, enabling resource-aware token requests during MCP authentication.
6. 0.38.1 Aug 2, 2026 · issue 001

   HolmesGPT 0.38.1 adds logging of the TOOL_MEMORY_LIMIT_MB remedy when a tool is OOM killed.

   └──▷ GET THIS VERSION
   copy

   $ git clone --branch 0.38.1 https://github.com/HolmesGPT/holmesgpt.git
   # already have the repo? check out this version:
   $ git checkout 0.38.1

   • ›Logs the TOOL_MEMORY_LIMIT_MB remedy when a tool is OOM killed, making memory-limit guidance visible in the investigation output.
7. 0.38.1 Aug 2, 2026 · issue -017

   HolmesGPT 0.38.1 adds logging of the TOOL_MEMORY_LIMIT_MB remedy when a tool is OOM killed.

   └──▷ GET THIS VERSION
   copy

   $ git clone --branch 0.38.1 https://github.com/HolmesGPT/holmesgpt.git
   # already have the repo? check out this version:
   $ git checkout 0.38.1

   • ›Logs the TOOL_MEMORY_LIMIT_MB remedy when a tool is OOM killed, making memory-limit guidance visible in the investigation output.
8. 0.38.0 Jul 30, 2026 · issue -020

   HolmesGPT 0.38.0 adds remote tool-call approval workflows and cuts prompt-token usage by 50–60%.

   └──▷ GET THIS VERSION
   copy

   $ git clone --branch 0.38.0 https://github.com/HolmesGPT/holmesgpt.git
   # already have the repo? check out this version:
   $ git checkout 0.38.0

   • ›Implements remote tool call approval workflows, enabling human-in-the-loop authorization of tool invocations from remote sessions.
   • ›Reduces system prompt and tool description sizes by ~50–60%, lowering prompt-token consumption and enabling use with tighter context-window limits.
   • ›Makes conversation-worker slot exhaustion visible and bounds reconnect sign-in attempts, improving observability of worker capacity limits.
   • ›Adds kubectl build to the Go binary distribution, expanding the toolchain's Kubernetes CLI surface.
9. 0.38.0 Jul 30, 2026 · issue 001

   HolmesGPT 0.38.0 adds remote tool-call approval workflows and cuts prompt token usage by 50–60%.

   └──▷ GET THIS VERSION
   copy

   $ git clone --branch 0.38.0 https://github.com/HolmesGPT/holmesgpt.git
   # already have the repo? check out this version:
   $ git checkout 0.38.0

   • ›Implements remote tool call approval workflows, enabling human-in-the-loop gating of agent-initiated tool executions.
   • ›Reduces system prompt and tool description sizes by ~50–60%, significantly cutting prompt token consumption per investigation.
   • ›Makes conversation-worker slot exhaustion visible and bounds reconnect sign-in attempts, improving observability of capacity limits.
   • ›Adds Slack channel-history context recovery evaluation coverage.
10. 0.37.0 Jul 27, 2026 · issue -023

    HolmesGPT 0.37.0 adds JSON log format for scrapers, Langfuse OpenTelemetry tracing, and AZURE_AD_TOKEN Entra ID auth support.

    └──▷ GET THIS VERSION
    copy

    $ git clone --branch 0.37.0 https://github.com/HolmesGPT/holmesgpt.git
    # already have the repo? check out this version:
    $ git checkout 0.37.0

    • ›Supports pre-acquired AZURE_AD_TOKEN environment variable for Entra ID authentication with Azure AD, enabling token-based auth flows without interactive login.
    • ›Adds JSON logging format support for log scrapers, enabling structured log ingestion pipelines.
    • ›Adds Langfuse support in OpenTelemetry tracing for LLM observability and trace analytics.
11. 0.37.0 Jul 27, 2026 · issue 001

    HolmesGPT 0.37.0 adds JSON log scraping, Langfuse/OpenTelemetry tracing, and pre-acquired AZURE_AD_TOKEN support for Entra ID auth.

    └──▷ GET THIS VERSION
    copy

    $ git clone --branch 0.37.0 https://github.com/HolmesGPT/holmesgpt.git
    # already have the repo? check out this version:
    $ git checkout 0.37.0

    • ›Supports pre-acquired AZURE_AD_TOKEN environment variable for Entra ID authentication, enabling token-based Azure auth flows without interactive login.
    • ›Adds JSON logging format support for log scrapers, enabling structured log ingestion pipelines.
    • ›Adds Langfuse support in OpenTelemetry tracing for LLM observability and trace analysis.
12. 0.36.0 Jul 13, 2026 · issue -037

    HolmesGPT 0.36.0 adds a /api/info endpoint, in-app TLS, Helm secret sugar, and multi-MCP-server tool-name isolation.

    └──▷ GET THIS VERSION
    copy

    $ git clone --branch 0.36.0 https://github.com/HolmesGPT/holmesgpt.git
    # already have the repo? check out this version:
    $ git checkout 0.36.0

    └──▷ TRY IT
    Check the running HolmesGPT instance's operational state — model, version, and configuration — without digging through logs.
    copy

    $ curl -s https://<holmesgpt-host>/api/info | jq .

    Mount an existing Kubernetes secret containing API keys into the HolmesGPT pod via Helm without manually listing each env var.
    yaml copy

    extraEnvVarsSecrets:
      - my-llm-api-keys

    • ›Adds GET /api/info endpoint for operational visibility into the running HolmesGPT instance.
    • ›Adds extraEnvVarsSecrets and envRef Helm values as shorthand for injecting secrets and configuring modelList without manual env var wiring.
    • ›Enables in-app HTTPS/TLS support for the HolmesGPT API, allowing TLS termination without an external proxy.
    • ›Supports multiple MCP servers simultaneously without tool-name collisions — identical tool names across servers are now disambiguated rather than overridden.
    • ›Adds a standalone manifest for provisioning additional Azure MCP accounts alongside existing ones.
    +5 moreshow less

    • ›Marks multi-instance-capable data sources in the data source catalog so operators can distinguish sources that support parallel deployments.
    • ›Consolidates Datadog and Elasticsearch/OpenSearch entries into unified catalog cards in the data source catalog.
    • ›Routes call_aws JSON-argument commands through the no-JSON path, cutting approximately 87 seconds of retries on AWS MCP calls.
    • ›GitHub MCP personal-access-token (PAT) path now uses native HTTP mode for more reliable authentication.
    • ›Bounds conversation and tool-call claiming to worker capacity, preventing oversubscription under load.
    └──▷ BREAKING ON UPGRADE

    • !KRR resource recommendation features have been removed from HolmesGPT.
13. 0.35.0 Jul 1, 2026 · issue -049

    HolmesGPT 0.35.0 adds env var templating in MCP OAuth URLs and expands the datasource catalog.

    └──▷ GET THIS VERSION
    copy

    $ git clone --branch 0.35.0 https://github.com/HolmesGPT/holmesgpt.git
    # already have the repo? check out this version:
    $ git checkout 0.35.0

    • ›Supports environment variable templating inside MCP OAuth URLs, enabling dynamic credential injection without hardcoding secrets in config.
    • ›Adds new data sources to the datasource catalog, broadening the range of backends HolmesGPT can query.
14. 0.34.0 Jun 28, 2026 · issue -052

    HolmesGPT 0.34.0 adds admin API endpoints for runtime config reload, event-driven operator triggers, and JWT-signed approval tickets for Kubernetes remediation.

    └──▷ GET THIS VERSION
    copy

    $ git clone --branch 0.34.0 https://github.com/HolmesGPT/holmesgpt.git
    # already have the repo? check out this version:
    $ git checkout 0.34.0

    • ›Adds admin API endpoints for runtime configuration reload, enabling live config changes without restarting the service.
    • ›Introduces event-driven triggers in operator mode, allowing HolmesGPT to react to Kubernetes events automatically.
    • ›Adds JWT-signed approval tickets to the Kubernetes Remediation MCP, with approval-based tool separation between safe and destructive actions.
    • ›Updates AWS MCP Helm addon to native Streamable HTTP (v2.1.0).
    • ›Adds firewall detection and user-friendly error handling for Supabase sign-in failures.
    +1 moreshow less

    • ›OAuth MCP now deletes expired token rows on 401 responses so users can re-authenticate cleanly.
    └──▷ BREAKING ON UPGRADE

    • !The deprecated Tool._is_restricted() mechanism has been removed; any toolset code relying on it will break.
15. 0.33.0 Jun 15, 2026 · issue -065

    HolmesGPT 0.33.0 adds cross-cluster remote tool execution and per-request tracing with span token metrics.

    └──▷ GET THIS VERSION
    copy

    $ git clone --branch 0.33.0 https://github.com/HolmesGPT/holmesgpt.git
    # already have the repo? check out this version:
    $ git checkout 0.33.0

    • ›Adds cross-cluster remote tool execution via an executor worker, toolset publishing, and caller headers, enabling tools to run against remote clusters from a central Holmes instance.
    • ›Adds per-request tracing experiment routing and span token metrics to the server for observability into LLM completion costs per request.
16. 0.32.0 Jun 11, 2026 · issue -069

    HolmesGPT 0.32.0 adds multi-instance toolsets, multi-cluster awareness, GitHub MCP auto-detection, and a unified firing boolean across all alert sources.

    └──▷ GET THIS VERSION
    copy

    $ git clone --branch 0.32.0 https://github.com/HolmesGPT/holmesgpt.git
    # already have the repo? check out this version:
    $ git checkout 0.32.0

    • ›Adds enable_todo flag to control the TodoWrite feature in evals.
    • ›Exposes a uniform firing boolean field for all alert sources, normalising alert state across integrations.
    • ›Adds multi-instance toolsets support ([ROB-3577]), allowing multiple instances of the same toolset to be configured simultaneously.
    • ›MCP auth health check now auto-detects and enables GitHub context automatically.
    • ›Adds a 'Denied commands' column to eval reports for visibility into blocked tool calls.
    +5 moreshow less

    • ›Introduces multi-cluster scope-awareness in the LLM prompt layer, enabling the AI to reason across clusters.
    • ›Hardwires cluster and conversation context into platform-MCP headers for richer contextual routing.
    • ›Supports overriding the config file at runtime.
    • ›Allows omitting the temperature parameter for models that reject it (e.g. Claude Opus 4.7+).
    • ›Marks the llm_summarize transformer as legacy / not recommended, steering users toward current alternatives.
17. 0.31.0 May 27, 2026 · issue -084

    HolmesGPT 0.31.0 adds Robusta Platform MCP toolset, multi-instance Elasticsearch/Grafana, per-conversation OAuth opt-out, and periodic MCP tool refresh.

    └──▷ GET THIS VERSION
    copy

    $ git clone --branch 0.31.0 https://github.com/HolmesGPT/holmesgpt.git
    # already have the repo? check out this version:
    $ git checkout 0.31.0

    • ›Adds the Robusta Platform MCP toolset with dynamic session token authentication for connecting Holmes to the Robusta platform.
    • ›Adds per-conversation OAuth opt-out support, letting individual sessions bypass OAuth flows.
    • ›Adds periodic MCP tools refresh so newly added tools are discovered without restarting the process.
    • ›Adds multi-instance Elasticsearch support, enabling Holmes to query across multiple Elasticsearch backends.
    • ›Adds multi-instance Grafana dashboard support, enabling Holmes to query across multiple Grafana instances.
    +4 moreshow less

    • ›Adds multi-region support for Robusta platform endpoints.
    • ›Adds optional ClickHouse HTTP JSONEachRow format for DateTime64 query results.
    • ›Adds evaluation harness for metrics analyses, expanding the benchmark coverage to metric-based investigations.
    • ›Warns at startup when an OTLP endpoint is configured but required OTEL packages are missing.
18. 0.30.1 May 20, 2026 · issue -091

    HolmesGPT 0.30.1 adds GitLab MCP datasource, user_email on chat requests, and LITELLM_MODEL_COST_MAP_URL support for firewalled environments.

    └──▷ GET THIS VERSION
    copy

    $ git clone --branch 0.30.1 https://github.com/HolmesGPT/holmesgpt.git
    # already have the repo? check out this version:
    $ git checkout 0.30.1

    • ›Accepts user_email on chat requests and persists it to HolmesUsageEvents for usage tracking.
    • ›Supports LITELLM_MODEL_COST_MAP_URL environment variable to configure model cost map fetching in firewalled environments.
    • ›Adds GitLab MCP entry to datasource-catalog.json, making GitLab available as a selectable MCP datasource.
    • ›Loki toolset now returns the raw response when parsing fails, surfacing unparseable data instead of silently dropping it.
19. 0.30.0 May 17, 2026 · issue -094

    HolmesGPT 0.30.0 adds GitLab MCP server integration alongside existing Kubernetes MCP support.

    └──▷ GET THIS VERSION
    copy

    $ git clone --branch 0.30.0 https://github.com/HolmesGPT/holmesgpt.git
    # already have the repo? check out this version:
    $ git checkout 0.30.0

    • ›Adds GitLab MCP server integration, enabling Holmes to interact with GitLab repositories and resources via the Model Context Protocol.
20. 0.29.0 May 14, 2026 · issue -097

    HolmesGPT 0.29.0 adds SSL/TLS custom CA support, AI usage tracking, Helm commonLabels, skills-from-Helm, and per-toolset prerequisite timeouts.

    └──▷ GET THIS VERSION
    copy

    $ git clone --branch 0.29.0 https://github.com/HolmesGPT/holmesgpt.git
    # already have the repo? check out this version:
    $ git checkout 0.29.0

    └──▷ USE IT
    Propagate custom labels (e.g. team, environment) to every resource the Holmes Helm chart creates.
    yaml copy

    commonLabels:
      team: security-ops
      environment: production

    Inject secrets or config from existing ConfigMaps/Secrets into the Holmes pod without hardcoding values.
    yaml copy

    additional_env_froms:
      - secretRef:
          name: my-llm-api-keys
      - configMapRef:
          name: holmes-extra-config

    • ›Adds commonLabels support to the Helm chart to propagate custom labels across all chart resources.
    • ›Adds additional_env_froms support to the Holmes Helm chart for injecting environment variables from ConfigMaps or Secrets.
    • ›Adds skills provisioning directly from Helm chart values.
    • ›Adds per-toolset prerequisite timeout to prevent startup hangs when toolset dependencies are unavailable.
    • ›Adds SSL/TLS support for custom CA bundles in the Realtime WebSocket connection.
    +1 moreshow less

    • ›Adds AI usage tracking via HolmesUsageEvents to monitor LLM consumption.
21. 0.28.0 May 7, 2026 · issue -104

    HolmesGPT 0.28.0 adds DNS configuration support and authenticated private-channel realtime subscriptions.

    └──▷ GET THIS VERSION
    copy

    $ git clone --branch 0.28.0 https://github.com/HolmesGPT/holmesgpt.git
    # already have the repo? check out this version:
    $ git checkout 0.28.0

    • ›Adds the ability to configure dnsPolicy and dnsConfig for HolmesGPT deployments.
    • ›Enables private channel support with authentication for realtime subscriptions.
22. 0.27.0 May 6, 2026 · issue -105

    HolmesGPT 0.27.0 adds GitHub Copilot AI provider, optional API key auth for HTTP endpoints, and a Kubernetes MCP addon.

    └──▷ GET THIS VERSION
    copy

    $ git clone --branch 0.27.0 https://github.com/HolmesGPT/holmesgpt.git
    # already have the repo? check out this version:
    $ git checkout 0.27.0

    • ›Adds optional API key authentication for all HTTP endpoints.
    • ›Adds GitHub Copilot as a supported AI provider.
    • ›Adds a Kubernetes MCP addon (ROB-3465 k8s mcp addon).
    • ›Adds support for additional environment variables in the Confluence MCP integration.
    • ›Adds support for editing tool command arguments before execution.
    +9 moreshow less

    • ›Adds configurable deployment rollout strategy via Helm.
    • ›Adds client secret support in the CLI OAuth flow, with an option to set the auth callback.
    • ›Adds an M2 Conversation Worker for async conversation processing, now enabled by default.
    • ›Adds health checks to detect silently-dead realtime WebSocket connections.
    • ›Adds Supabase Realtime verification before starting the conversation worker.
    • ›Adds retry logic to the conversation worker.
    • ›Extracts text from MCP resource content blocks in tool results.
    • ›Injects frontend tools into the worker conversation executor.
    • ›Disables todowrite behavior controls in scheduled prompts.
23. 0.26.0 Apr 28, 2026 · issue -113

    HolmesGPT 0.26.0 introduces Skills extensibility, VictoriaLogs toolset, and new Grafana timeout controls

    └──▷ GET THIS VERSION
    copy

    $ git clone --branch 0.26.0 https://github.com/HolmesGPT/holmesgpt.git
    # already have the repo? check out this version:
    $ git checkout 0.26.0

    • ›Adds extraVolumes, extraVolumeMounts, and extraInitContainers configuration keys to the AWS MCP toolset for flexible volume and init-container injection.
    • ›Adds max retries and timeout configuration for Grafana integrations.
    • ›Introduces Skills, a new extensibility model enabling custom tool integrations and workflows via configuration.
    • ›Adds a VictoriaLogs toolset for log search and analysis.
    • ›Adds support for logging aggregated investigation token usage.
    +2 moreshow less

    • ›Improves the data source configuration flow.
    • ›Adds Atlassian admin setup steps to OAuth MCP documentation, expanding supported MCP OAuth configuration guidance.
    └──▷ BREAKING ON UPGRADE

    • !The Skills functionality introduces a new extensibility model — existing custom tool integrations and workflows may need to be migrated to the new Skills configuration format.
24. 0.25.0 Apr 23, 2026 · issue -118

    HolmesGPT 0.25.0 adds a Crossplane toolset and OAuth support for MCP servers in CLI and server mode.

    └──▷ GET THIS VERSION
    copy

    $ git clone --branch 0.25.0 https://github.com/HolmesGPT/holmesgpt.git
    # already have the repo? check out this version:
    $ git checkout 0.25.0

    • ›Adds OAuth support for MCP servers in both CLI and server mode.
    • ›Adds a Crossplane toolset for investigating Crossplane resources.
25. 0.24.4 Apr 20, 2026 · issue -121

    HolmesGPT 0.24.4 adds OpenTelemetry instrumentation, YAML anchors in config, new Kafka tools, and expanded DB data source support.

    └──▷ GET THIS VERSION
    copy

    $ git clone --branch 0.24.4 https://github.com/HolmesGPT/holmesgpt.git
    # already have the repo? check out this version:
    $ git checkout 0.24.4

    └──▷ USE IT
    Reuse shared config blocks (e.g., common LLM settings) across multiple toolsets without repetition.
    yaml copy

    anchors:
      common_llm: &common_llm
        model: gpt-4o
        max_tokens: 4096
    
    toolsets:
      - name: my-toolset
        llm:
          <<: *common_llm

    • ›Adds anchors top-level field to HolmesGPT config to support YAML anchors, enabling reuse of repeated config blocks.
    • ›Adds OpenTelemetry tracing and metrics instrumentation for observing HolmesGPT's own runtime behavior.
    • ›Adds missing Kafka tools to match the full capability set of the kafka-mcp-server integration.
    • ›Expands datasource support with new database data sources and improved datasource descriptions.
    • ›Adds Zabbix integration with supporting documentation.
    +2 moreshow less

    • ›Allows using the table API without specifying a version.
    • ›Rebrands Azure OpenAI integration to Azure AI Foundry.
26. 0.24.1 Apr 7, 2026 · issue -134

    HolmesGPT 0.24.1 adds mTLS/SSL support to the Kafka/admin toolset.

    └──▷ GET THIS VERSION
    copy

    $ git clone --branch 0.24.1 https://github.com/HolmesGPT/holmesgpt.git
    # already have the repo? check out this version:
    $ git checkout 0.24.1

    • ›Adds mTLS/SSL support to the Kafka/admin toolset, enabling encrypted and mutually authenticated connections to Kafka brokers.
27. 0.24.0 Apr 5, 2026 · issue -136

    HolmesGPT 0.24.0 adds frontend tool execution, a startup sample-questions menu, and configurable Helm security contexts.

    └──▷ GET THIS VERSION
    copy

    $ git clone --branch 0.24.0 https://github.com/HolmesGPT/holmesgpt.git
    # already have the repo? check out this version:
    $ git checkout 0.24.0

    └──▷ TRY IT
    Ask Holmes a question without specifying a subcommand — useful for quick ad-hoc investigations from the shell.
    copy

    $ holmesgpt 'Why is my payment-service pod crash-looping?'

    • ›Adds configurable podSecurityContext and containerSecurityContext fields to the Holmes Helm chart, enabling custom pod and container security contexts at deploy time.
    • ›Adds frontend tool execution support to Holmes, allowing UI-side tools to be invoked through the HTTP API (documented in the HTTP API frontend tools implementation guide).
    • ›Defaults to the ask subcommand when no subcommand is provided on the CLI, so bare holmesgpt <question> invocations now work without specifying a subcommand.
    • ›Adds a sample questions menu on startup to guide practitioners toward common investigations.
28. 0.23.0 Mar 26, 2026 · issue -146

    HolmesGPT 0.23.0 adds a revamped CLI experience, multi-account NewRelic support, image content in MCP results, and HTTP Basic Auth for ServiceNow.

    └──▷ GET THIS VERSION
    copy

    $ git clone --branch 0.23.0 https://github.com/HolmesGPT/holmesgpt.git
    # already have the repo? check out this version:
    $ git checkout 0.23.0

    • ›Adds HTTP Basic Auth support to the ServiceNow tables toolset.
    • ›Adds support for image content in MCP tool results.
    • ›Adds support for multiple NewRelic accounts in a single HolmesGPT instance.
    • ›Adds a Docker Compose setup for simplified HolmesGPT deployment.
    • ›Delivers a new and improved interactive CLI experience.
29. 0.22.0 Mar 18, 2026 · issue -154

    HolmesGPT 0.22.0 adds MongoDB toolset, mTLS for HTTP/Elasticsearch, digest auth, Inspektor Gadget filtering, and GitHub App token management.

    └──▷ GET THIS VERSION
    copy

    $ git clone --branch 0.22.0 https://github.com/HolmesGPT/holmesgpt.git
    # already have the repo? check out this version:
    $ git checkout 0.22.0

    └──▷ TRY IT
    Use a separate classifier model by setting the CLASSIFIER_MODEL env var, e.g. when routing classification to a cheaper model via OpenRouter.
    copy

    $ export CLASSIFIER_MODEL=openrouter/anthropic/claude-3-haiku
    holmes ask 'Why is my pod crashlooping?'

    • ›Adds initContainers and args support to the Helm chart for advanced pod configuration.
    • ›Adds digest auth and mTLS support to the HTTP toolset, configurable via the HTTP connector config.
    • ›Adds mTLS (mutual TLS) support to the Elasticsearch toolset.
    • ›Adds CLASSIFIER_MODEL environment variable support for specifying a separate classifier model (documented alongside OpenRouter requirements).
    • ›Adds behavior_controls API parameter for prompt customization in the investigation API.
    +13 moreshow less

    • ›Adds tool_results_dir parameter to ToolCallingLLM in the custom LLM interface.
    • ›Adds a MongoDB toolset for querying and diagnostics.
    • ›Adds filtering support to the Inspektor Gadget toolset, enabling scoped tool invocations.
    • ›Adds GitHub App token manager for automatic installation token handling, including MCP config support for GitHub App credentials.
    • ›Adds datasource-catalog.json for datasource cataloging.
    • ›Adds keyless auth support for Azure AI Foundry.
    • ›Adds cached tokens tracking and max_prompt_tokens_per_call tracking to LLM usage/cost reporting.
    • ›Adds URL query parameter support to tab synchronization in the TUI.
    • ›Adds Enum field support and MCP config management to the TUI.
    • ›Adds multi-round approval workflow support with iteration offset.
    • ›Enables strict tool calling universally with per-tool compatibility checks.
    • ›Disables unconfigured toolsets by default in the CLI.
    • ›Removes the automated investigation API, consolidating all investigation flows into the conversational interface.
    └──▷ BREAKING ON UPGRADE

    • !The automated investigation API has been removed; all investigation flows must now use the conversational interface.
    • !Unconfigured toolsets are now disabled by default in the CLI — toolsets that were previously active without explicit configuration will no longer run.
    • !Message truncation logic has been removed; context overflow now causes a fast failure instead of silently truncating.
30. 0.21.0 Mar 6, 2026 · issue -165

    HolmesGPT 0.21.0 adds Coralogix archive support, Confluence integration, an interactive TUI for toolset config, and AWS Bedrock bearer-token auth.

    └──▷ GET THIS VERSION
    copy

    $ git clone --branch 0.21.0 https://github.com/HolmesGPT/holmesgpt.git
    # already have the repo? check out this version:
    $ git checkout 0.21.0

    • ›Adds AWS_BEARER_TOKEN_BEDROCK environment variable to authenticate with AWS Bedrock models.
    • ›Adds an interactive TUI for toolset configuration, letting CLI users configure toolset prerequisites without manual file editing.
    • ›Adds Coralogix archive support as a new data-source integration.
    • ›Adds Confluence as a new toolset integration for knowledge-base lookups during investigations.
    • ›Adds a direct Prometheus endpoint option for Grafana Cloud data sources.
    +7 moreshow less

    • ›Supports recursive parsing of nested schemas in MCP tool parameters, enabling richer MCP toolset interoperability.
    • ›Adds HTTP header propagation to forward request headers through HolmesGPT's internal calls.
    • ›Adds resourcequotas RBAC role to the Holmes default service account, expanding Kubernetes resource visibility.
    • ›Adds webhook configuration permissions to the HolmesGPT service account.
    • ›Adds appProtocol field to the Holmes Helm service port configuration.
    • ›Implements lazy initialization for toolset prerequisites in CLI mode, so missing prerequisites no longer block unrelated toolsets at startup.
    • ›Adds Claude Opus 4.6 and Sonnet 4.6 to benchmark model defaults.
31. 0.20.0 Feb 24, 2026 · issue -174

    HolmesGPT 0.20.0 adds SQL database toolsets, five MCP server integrations, Inspektor Gadget, KubeVela, and a generic HTTP connector toolset.

    └──▷ GET THIS VERSION
    copy

    $ git clone --branch 0.20.0 https://github.com/HolmesGPT/holmesgpt.git
    # already have the repo? check out this version:
    $ git checkout 0.20.0

    • ›Adds SQL Database toolsets covering PostgreSQL, ClickHouse, MariaDB, MySQL, SQLServer, and SQLite, enabling AI-driven investigation of database state.
    • ›Adds a generic HTTP connector toolset for querying arbitrary HTTP endpoints as investigation sources.
    • ›Adds Inspektor Gadget toolset for kernel-level eBPF-based diagnostics within HolmesGPT investigations.
    • ›Adds KubeVela built-in toolsets for investigating OAM application workloads.
    • ›Adds Grafana MCP server integration, enabling HolmesGPT to query Grafana via the Model Context Protocol.
    +8 moreshow less

    • ›Adds Sentry MCP server integration for pulling error and event context from Sentry during investigations.
    • ›Adds Prefect MCP server integration for workflow observability via the Model Context Protocol.
    • ›Adds Confluence MCP server integration for surfacing runbook and knowledge-base content during investigations.
    • ›Adds Kubernetes Remediation MCP integration, allowing HolmesGPT to execute remediation actions against Kubernetes clusters.
    • ›Adds filesystem storage for large tool results, removing in-memory size constraints on investigation outputs.
    • ›Adds multi-cluster support and external change handling to Robusta tools.
    • ›Adds GCP workload identity support for the GCP MCP integration.
    • ›Adds OpenRouter as a supported AI provider.
    └──▷ BREAKING ON UPGRADE

    • !Additional instructions support has been removed; any configuration relying on that feature will no longer work after upgrading.
    • !Unused Coralogix label settings have been removed and related functions have been flattened; configurations referencing those Coralogix label settings will break on upgrade.
32. 0.19.0 Jan 27, 2026 · issue -202

    HolmesGPT 0.19.0 adds GitHub, Confluence, and Splunk MCP servers, image input support, and scheduled prompt investigations.

    └──▷ GET THIS VERSION
    copy

    $ git clone --branch 0.19.0 https://github.com/HolmesGPT/holmesgpt.git
    # already have the repo? check out this version:
    $ git checkout 0.19.0

    • ›Adds Splunk MCP server support, replacing any prior ad-hoc Splunk integration with an explicit MCP-based server.
    • ›Adds GitHub MCP server, replacing the legacy GitHub toolset with an MCP-based implementation.
    • ›Adds Confluence MCP server for Confluence-based investigations and workflows.
    • ›Supports passing an auth token to MCP servers via HTTP header.
    • ›Adds support for image input to the chat API, including a new images field in the chat API.
    +3 moreshow less

    • ›Adds daily and weekly scheduled prompt investigations in SAS.
    • ›Redesigns the Bash toolset with prefix-based approvals for command execution control.
    • ›Improves OOMKill detection capability.
    └──▷ BREAKING ON UPGRADE

    • !The git toolset has been deprecated in favor of the Git MCP server; existing configurations relying on the git toolset will stop working and must be migrated to the MCP-based implementation.
33. 0.18.5 Jan 19, 2026 · issue -210

    HolmesGPT 0.18.5 adds structured output to /api/chat, periodic toolset refresh, and Claude tool-approval support

    └──▷ GET THIS VERSION
    copy

    $ git clone --branch 0.18.5 https://github.com/HolmesGPT/holmesgpt.git
    # already have the repo? check out this version:
    $ git checkout 0.18.5

    • ›Adds structured output support to the POST /api/chat endpoint, enabling callers to receive machine-parseable responses.
    • ›Adds periodic toolset status refresh in server mode, keeping tool availability current without restarting the process.
    • ›Adds Claude tool-approval support, allowing all tools to be approved for Claude-backed workflows.
    • ›Adds SSL verification disable option for Claude integrations.
    • ›Adds long file name support on Windows, removing path-length errors when processing deeply nested files.
    +1 moreshow less

    • ›Helm chart now uses a dynamic imagePullPolicy driven by values, replacing the previously hardcoded policy.
34. 0.18.4 Jan 15, 2026 · issue -214

    HolmesGPT 0.18.4 adds Azure managed Prometheus, New Relic multi-account, GitHub MCP, and AWS MCP improvements.

    └──▷ GET THIS VERSION
    copy

    $ git clone --branch 0.18.4 https://github.com/HolmesGPT/holmesgpt.git
    # already have the repo? check out this version:
    $ git checkout 0.18.4

    • ›Adds Azure managed Prometheus support to the Prometheus toolset.
    • ›Adds New Relic multiple-account support, enabling queries across more than one New Relic account.
    • ›Adds GitHub MCP addon for the Helm chart, extending MCP integrations alongside the existing Azure MCP toolset.
    • ›Improves AWS MCP integration with several enhancements.
    • ›Enables the connectivity_check toolset by default in the Helm chart.
    +2 moreshow less

    • ›Adds support for an additional system prompt URL in eval configurations.
    • ›Improves SSL error messaging for the Prometheus toolset to surface clearer diagnostics.
35. 0.18.2 Jan 6, 2026 · issue -223

    HolmesGPT 0.18.2 adds TCP connectivity checks, GCP & AKS MCP integrations, and an improved Elasticsearch/OpenSearch toolset.

    └──▷ GET THIS VERSION
    copy

    $ git clone --branch 0.18.2 https://github.com/HolmesGPT/holmesgpt.git
    # already have the repo? check out this version:
    $ git checkout 0.18.2

    • ›Adds a connectivity check toolset supporting TCP checks for network reachability diagnostics.
    • ›Adds GCP MCP integration for Google Cloud Platform context in investigations.
    • ›Adds Azure (AKS) MCP integration for Azure Kubernetes Service context.
    • ›Updates the Prometheus toolset configuration.
    └──▷ BREAKING ON UPGRADE

    • !The Elasticsearch/OpenSearch toolset has breaking configuration changes — consult the updated docs at https://holmesgpt.dev/data-sources/builtin-toolsets/elasticsearch/#elasticsearch-opensearch before upgrading.
36. 0.18.1 Dec 28, 2025 · issue -232

    HolmesGPT 0.18.1 adds memory usage limits for bash and yaml toolsets via ulimit.

    └──▷ GET THIS VERSION
    copy

    $ git clone --branch 0.18.1 https://github.com/HolmesGPT/holmesgpt.git
    # already have the repo? check out this version:
    $ git checkout 0.18.1

    • ›Adds ulimit-based memory limiting for the bash and yaml toolsets to cap resource consumption during automated investigations.
37. 0.18.0 Dec 26, 2025 · issue -234

    HolmesGPT 0.18.0 adds runbook support by alerts and symptoms, external links in tool results, and new health endpoints for pod probes.

    └──▷ GET THIS VERSION
    copy

    $ git clone --branch 0.18.0 https://github.com/HolmesGPT/holmesgpt.git
    # already have the repo? check out this version:
    $ git checkout 0.18.0

    • ›Adds healthz and readyz endpoints for pod probe health checks.
    • ›Adds runbook alerts to the LLM prompt, enabling runbook lookup by alerts and symptoms.
    • ›Tool results now include external links for quick navigation to source resources.
    • ›Reduces LLM context consumption by serializing multiple tool call outputs as minified JSON.
    • ›Improves Coralogix toolset with expanded capabilities.
    +4 moreshow less

    • ›Adds health checks to TCP connections.
    • ›Publishes Helm charts via OCI registry.
    • ›Expands default CRD permissions for broader Kubernetes resource coverage.
    • ›Adds timeout to Kafka prerequisites check.
    └──▷ BREAKING ON UPGRADE

    • !The datadog/rds plugin is removed; its functionality is now covered by the datadog/metrics plugin.
38. 0.17.0 Dec 9, 2025 · issue -251

    HolmesGPT 0.17.0 adds MariaDB MCP addon, user runbooks catalog, and total cost display with --log-costs.

    └──▷ GET THIS VERSION
    copy

    $ git clone --branch 0.17.0 https://github.com/HolmesGPT/holmesgpt.git
    # already have the repo? check out this version:
    $ git checkout 0.17.0

    └──▷ TRY IT
    Track the total LLM spend for a Holmes investigation to understand cost per incident.
    copy

    $ holmes ask 'Why is my payment-service pod crash-looping?' --log-costs

    • ›Adds --log-costs flag now displays total cost at the end of a run, in addition to per-call costs.
    • ›Adds MariaDB MCP addon, extending the set of supported database integrations.
    • ›Adds support for a user runbooks catalog, enabling practitioners to supply their own runbooks for Holmes to reference during investigations.
39. 0.16.4 Dec 1, 2025 · issue -259

    HolmesGPT 0.16.4 adds stdio MCP support and AWS MCP instructions.

    └──▷ GET THIS VERSION
    copy

    $ git clone --branch 0.16.4 https://github.com/HolmesGPT/holmesgpt.git
    # already have the repo? check out this version:
    $ git checkout 0.16.4

    • ›Adds stdio transport support for MCP (Model Context Protocol) servers, enabling local process-based tool integrations.
    • ›Adds AWS MCP instructions, enabling Holmes to interact with AWS resources via the MCP toolchain.
40. 0.16.3 Nov 23, 2025 · issue -267

    HolmesGPT 0.16.3 adds ServiceNow and OpenShift toolsets, streamable-HTTP MCP support, and Loki improvements.

    └──▷ GET THIS VERSION
    copy

    $ git clone --branch 0.16.3 https://github.com/HolmesGPT/holmesgpt.git
    # already have the repo? check out this version:
    $ git checkout 0.16.3

    └──▷ TRY IT
    Cap how many tokens a single tool can consume from the context window, useful when running against models with tight context limits.
    copy

    $ export TOOL_MAX_ALLOCATED_CONTEXT_WINDOW_TOKENS=8000
    holmes ask 'Why is my payment-service pod crash-looping?'

    • ›Adds TOOL_MAX_ALLOCATED_CONTEXT_WINDOW_TOKENS environment variable to control the maximum context window tokens allocated per tool.
    • ›New ServiceNow toolset for querying and interacting with ServiceNow from within HolmesGPT investigations.
    • ›New OpenShift toolset enabling use of oc commands during AI-driven investigations.
    • ›Adds streamable-HTTP transport support for MCP (Model Context Protocol) integrations.
    • ›Supports custom GitHub Enterprise URLs, allowing users on GitHub Enterprise to point HolmesGPT at their own instance.
    +2 moreshow less

    • ›Improved Loki toolset with expanded log querying capabilities.
    • ›Improved KRR (Kubernetes Resource Recommender) tool integration.
41. 0.16.1 Nov 3, 2025 · issue -287

    HolmesGPT 0.16.1 adds Grafana dashboard tools, CRD reading, Robusta runbook support, and a resource-issues fetch tool.

    └──▷ GET THIS VERSION
    copy

    $ git clone --branch 0.16.1 https://github.com/HolmesGPT/holmesgpt.git
    # already have the repo? check out this version:
    $ git checkout 0.16.1

    • ›Adds four new built-in tools for retrieving Grafana dashboards, enabling Holmes to pull and reason over dashboard data during investigations.
    • ›Enables Holmes to read Kubernetes Custom Resource Definitions (CRDs), expanding its cluster-resource awareness beyond core API objects.
    • ›Adds a fetch-resource-issues tool that retrieves workload issues directly, supporting richer automated triage workflows.
    • ›Adds Robusta runbook support, allowing Holmes to reference and execute Robusta runbooks during incident response.
    • ›Adds support for pulling external changes, keeping Holmes tooling in sync with upstream configurations.
    +1 moreshow less

    • ›Improves prompt caching to reduce latency and cost on repeated or similar queries.
42. 0.16.0 Oct 21, 2025 · issue -300

    HolmesGPT 0.16.0 adds AWS MCP integration for IAM/EC2/cost queries and conversation history compaction for longer sessions.

    └──▷ GET THIS VERSION
    copy

    $ git clone --branch 0.16.0 https://github.com/HolmesGPT/holmesgpt.git
    # already have the repo? check out this version:
    $ git checkout 0.16.0

    • ›New AWS MCP addon enables querying IAM Roles, security groups, EC2 instances, and AWS costs directly from HolmesGPT conversations.
    • ›Adds conversation history compaction to support longer, data-heavy conversations without hitting context limits.
    • ›Adds AI reasoning logs for visibility into the model's decision process during investigations.
43. 0.15.0 Oct 13, 2025 · issue -308

    HolmesGPT 0.15.0 adds Cilium/Hubble toolsets, a feedback command, AG-UI chat endpoint, and a tool approval workflow.

    └──▷ GET THIS VERSION
    copy

    $ git clone --branch 0.15.0 https://github.com/HolmesGPT/holmesgpt.git
    # already have the repo? check out this version:
    $ git checkout 0.15.0

    └──▷ TRY IT
    Submit feedback directly from the CLI after an investigation session.
    copy

    $ holmes feedback

    • ›Adds feedback subcommand (with privacy notice) for submitting feedback from the CLI.
    • ›Allows passing a custom system-prompt parameter in the Holmes chat endpoint via the custom_sys_prompt param (ROB-2197).
    • ›Adds experimental AG-UI-supported chat endpoint and PPL query-assist feature.
    • ›Adds Cilium and Hubble toolsets for network observability and policy inspection.
    • ›Adds a tool approval workflow, enabling operator review of tool calls before execution (ROB-1933).
    +7 moreshow less

    • ›Adds Vertex AI Claude model support by including the google-cloud-aiplatform dependency.
    • ›Adds OpenSearchQueryAssistToolset with prerequisites checking.
    • ›Expands NewRelic toolset with freeform queries, transaction embed support, and improved trace handling.
    • ›Limits max tool output to a hard-coded 25K tokens regardless of model context size, improving stability with large responses.
    • ›Improves Gemini model support and compatibility.
    • ›Removes experimental tag from three previously experimental toolsets.
    • ›Improves CLI output to reduce noise during normal operation.
44. 0.14.2 Sep 16, 2025 · issue -334

    HolmesGPT 0.14.2 adds improved Tempo integration, Claude Sonnet 1M token model support, and context window optimizations for metrics and trace data.

    └──▷ GET THIS VERSION
    copy

    $ git clone --branch 0.14.2 https://github.com/HolmesGPT/holmesgpt.git
    # already have the repo? check out this version:
    $ git checkout 0.14.2

    • ›Supports the Sonnet 1M token model for extended context analysis.
    • ›Improves Tempo integration to enable latency increase and HTTP 500 error root-cause queries.
    • ›Optimizes context window handling for metrics and trace data, increasing answer accuracy.
    • ›Improves Datadog integration accuracy through multiple small changes.
45. 0.14.1 Sep 15, 2025 · issue -335

    HolmesGPT 0.14.1 adds a free-form Tempo toolset, LLM-based message summarization via a Transformer mechanism, and api_base/api_version Azure OpenAI config keys.

    └──▷ GET THIS VERSION
    copy

    $ git clone --branch 0.14.1 https://github.com/HolmesGPT/holmesgpt.git
    # already have the repo? check out this version:
    $ git checkout 0.14.1

    • ›Adds api_base and api_version config keys for Azure OpenAI support in the HolmesGPT configuration.
    • ›Introduces LLM-based message summarization via a new Transformer mechanism to optimize context window handling, improving accuracy with metrics and trace data.
    • ›Adds a free-form Tempo toolset enabling natural-language trace queries such as 'Why was there an increase in latency?' and 'Find the cause of recent HTTP 500 errors.'
    • ›Implements an LLM Model Registry allowing context size overrides per model.
    • ›Adds max_tokens and max output tokens fields to Holmes responses.
    +4 moreshow less

    • ›Respects DEFAULT_LOG_LIMIT in the Loki integration for more controlled log retrieval.
    • ›Improves Prometheus integration with better truncation and a reduced max query size cap.
    • ›Improves DataDog integration with accuracy and error-handling enhancements.
    • ›Improves Ollama support.
46. 0.13.2 Sep 3, 2025 · issue -347

    HolmesGPT 0.13.2 adds Datadog integration, cross-account IRSA, Prometheus range queries, and CLI approval for bash tool execution.

    └──▷ GET THIS VERSION
    copy

    $ git clone --branch 0.13.2 https://github.com/HolmesGPT/holmesgpt.git
    # already have the repo? check out this version:
    $ git checkout 0.13.2

    • ›Adds a general-purpose Datadog tool, enabling HolmesGPT to query Datadog as part of AI-driven investigations.
    • ›Adds cross-account IRSA support, allowing HolmesGPT to assume IAM roles across AWS accounts for multi-account Kubernetes environments.
    • ›Adds Prometheus range-query support to the Prometheus toolset, enabling historical metric analysis over configurable time windows.
    • ›Adds a CLI approval step for the bash tool, requiring explicit user confirmation before executing bash commands from the CLI.
    • ›Adds manual and HPA-driven replica count control options.
47. 0.13.1 Aug 27, 2025 · issue -354

    HolmesGPT 0.13.1 adds AWS Bedrock single-token auth and AMP IRSA support for IAM-based authentication.

    └──▷ GET THIS VERSION
    copy

    $ git clone --branch 0.13.1 https://github.com/HolmesGPT/holmesgpt.git
    # already have the repo? check out this version:
    $ git checkout 0.13.1

    • ›Adds single-token authentication support for AWS Bedrock as an LLM backend.
    • ›Adds IRSA (IAM Roles for Service Accounts) support for Amazon Managed Prometheus (AMP) integration.
    • ›Replaces CLI toolsets with a bash toolset.
48. 0.13.0 Aug 20, 2025 · issue -361

    HolmesGPT 0.13.0 adds a Datadog RDS analysis toolset, REASONING_EFFORT env var, streaming API chat, custom annotations, and subtask support.

    └──▷ GET THIS VERSION
    copy

    $ git clone --branch 0.13.0 https://github.com/HolmesGPT/holmesgpt.git
    # already have the repo? check out this version:
    $ git checkout 0.13.0

    • ›Adds support for the REASONING_EFFORT environment variable to control LLM reasoning effort at runtime.
    • ›Adds call_stream support for the API chat endpoint, enabling streaming responses from the Holmes API.
    • ›Adds an option for custom annotations on Holmes resources.
    • ›Adds a new Datadog RDS analysis toolset for database-focused investigations via Datadog.
    • ›Adds subtask support, allowing Holmes investigations to spawn and track sub-investigations.
    +7 moreshow less

    • ›Adds intermediate AI message and reasoning output via ai_message in investigation responses.
    • ›Adds LLM support to the investigation workflow, broadening which models can drive investigations.
    • ›Adds Prometheus AMP (Amazon Managed Prometheus) support alongside fixes to loading configs from environment variables.
    • ›Adds icon_url field to tool call results in the API response.
    • ›Adds revisionhistorylimit support for Kubernetes deployment configuration.
    • ›Returns rate-limit errors to callers when the upstream LLM rate-limits Holmes.
    • ›Adds an AI safety prompt to the system prompt for all Holmes interactions.
49. 0.12.4 Aug 4, 2025 · issue -363

    HolmesGPT 0.12.4 adds MCP server config file support, configurable config dir, OpenShift Prometheus, YAML syntax highlighting, and faster startup.

    └──▷ GET THIS VERSION
    copy

    $ git clone --branch 0.12.4 https://github.com/HolmesGPT/holmesgpt.git
    # already have the repo? check out this version:
    $ git checkout 0.12.4

    └──▷ USE IT
    Configure an MCP server declaratively in the HolmesGPT config file instead of passing CLI flags each run.
    yaml copy

    mcp_servers:
      - name: my-mcp-server
        url: http://localhost:8080

    Point HolmesGPT at a non-default config directory, useful in multi-tenant or containerized environments.
    copy

    $ holmes ask --config-dir /etc/holmesgpt/prod 'Why is my pod crash-looping?'

    • ›Adds support for configuring MCP server in the config file, enabling declarative MCP server setup without CLI flags.
    • ›Makes the config file directory configurable, allowing non-default config locations to be specified.
    • ›Adds a toggle for the version check, letting users enable or disable the startup version check.
    • ›Makes the agent name in the welcome banner configurable.
    • ›Adds autocomplete for the /show command in interactive mode.
    +9 moreshow less

    • ›Adds YAML syntax highlighting for output in the /show command.
    • ›Adds OpenShift Prometheus support as a metrics toolset.
    • ›Enables HolmesGPT to unzip logs from findings, expanding log analysis coverage.
    • ›Adds cluster name to the system prompt, giving the LLM context about which cluster it is operating on.
    • ›Adds line count to tool output for better readability of large results.
    • ›Improves startup speed of holmes ask via lazy imports.
    • ›Requests multiple tool calls for reasoning models to improve accuracy.
    • ›Improves HolmesGPT accuracy on questions about itself, configuring tools, and using runbooks.
    • ›Improves DataDog metrics toolset.
50. 0.12.3 Jul 24, 2025 · issue -364

    HolmesGPT 0.12.3 adds Datadog, Azure SQL, ServiceNow, Atlas MongoDB, and NSG toolsets plus interactive slash commands and a --refresh-toolsets flag.

    └──▷ GET THIS VERSION
    copy

    $ git clone --branch 0.12.3 https://github.com/HolmesGPT/holmesgpt.git
    # already have the repo? check out this version:
    $ git checkout 0.12.3

    └──▷ TRY IT
    Force a fresh toolset status check when a toolset may have become available since the last run.
    copy

    $ holmes ask --refresh-toolsets 'Why is my pod crash-looping?'

    Pipe kubectl output directly into Holmes for inline investigation without a separate ask step.
    copy

    $ kubectl describe pod my-app-51 | holmes ask 'What is wrong with this pod?'

    • ›Adds --refresh-toolsets flag to the ask command to force a refresh of toolset status on demand.
    • ›Introduces /run, /clear, /show, and /context slash commands in interactive mode.
    • ›Adds Datadog logs toolset, Datadog metrics toolset, and Datadog traces toolset for querying logs, metrics, and traces from Datadog.
    • ›Adds Azure SQL toolset for querying and troubleshooting Azure SQL databases.
    • ›Adds Azure Network Security Groups (NSGs) toolset for inspecting network security group rules.
    +14 moreshow less

    • ›Adds Atlas MongoDB toolset for querying MongoDB Atlas clusters.
    • ›Adds ServiceNow experimental toolset for integrating with ServiceNow.
    • ›Introduces a runbook toolset (feat(runbook)) to fetch and apply internal runbooks during investigations.
    • ›Enriches AKS node health tools with additional diagnostic capabilities.
    • ›Adds a toolset management tool command and caches toolset status to speed up repeated invocations.
    • ›Enables Holmes to count Kubernetes resources during investigations.
    • ›Makes interactive mode the default when running the CLI.
    • ›Adds support for piped input to the Holmes CLI.
    • ›Makes OpenSearch environment variable and config description more generic for broader compatibility.
    • ›Adds tracing support to the Holmes CLI.
    • ›Injects global date context into prompts so the LLM is aware of the current date.
    • ›Checks Prometheus toolset health during auto-discovery, not only on explicit invocation.
    • ›Checks toolset prerequisites in parallel, reducing CLI startup time.
    • ›Adds an HTTP API with published documentation.
51. 0.11.5 Jun 5, 2025 · issue -365

    HolmesGPT 0.11.5 adds Remote MCP server support for real-time access to external data sources and tools.

    └──▷ GET THIS VERSION
    copy

    $ git clone --branch 0.11.5 https://github.com/HolmesGPT/holmesgpt.git
    # already have the repo? check out this version:
    $ git checkout 0.11.5

    └──▷ HOW TO FIND IT
    Connect HolmesGPT to a remote MCP server to give it real-time access to an external tool or data source during investigations.
    📍See https://docs.robusta.dev/master/configuration/holmesgpt/remote_mcp_servers.html for the full configuration reference.
    • ›Adds Remote MCP server support, enabling HolmesGPT to connect to external data sources and tools in real time via configured remote endpoints.
    • ›Loads the Robusta UI token from an environment variable when required, reducing manual credential configuration.
    • ›Allows overriding the Grafana health-check URL via configuration, enabling custom Grafana deployment paths.
    • ›Adds Azure OpenAI to the list of supported models.
    • ›Removes the mandatory requirement for resource group, subscription, and cluster name in the AKS toolset, simplifying Azure configuration.
    +1 moreshow less

    • ›Restarts the Holmes pod automatically on config changes, ensuring configuration updates are applied without manual intervention.
52. 0.11.2 May 12, 2025 · issue -366

    HolmesGPT 0.11.2 adds direct connection support for Loki and Tempo datasources.

    └──▷ GET THIS VERSION
    copy

    $ git clone --branch 0.11.2 https://github.com/HolmesGPT/holmesgpt.git
    # already have the repo? check out this version:
    $ git checkout 0.11.2

    • ›Supports direct connections to Loki and Tempo, removing the requirement to route through a Grafana datasource.
53. 0.11.1 May 12, 2025 · issue -366

    HolmesGPT 0.11.1 adds RabbitMQ, Elasticsearch, AKS, and Git toolsets plus Prometheus auto-discovery and multi-tenant Loki/Grafana support.

    └──▷ GET THIS VERSION
    copy

    $ git clone --branch 0.11.1 https://github.com/HolmesGPT/holmesgpt.git
    # already have the repo? check out this version:
    $ git checkout 0.11.1

    └──▷ TRY IT
    Supply a saved investigation prompt from a file instead of typing it inline — useful for repeatable runbook-style queries in CI or on-call workflows.
    copy

    $ holmes ask --prompt-file ./prompts/oom-investigation.txt

    Use a named AWS profile when invoking Holmes with a Bedrock-backed model, avoiding the need to export credentials directly.
    copy

    $ AWS_PROFILE=prod-sre holmes ask 'Why is the payment service crashing?'

    • ›Adds --prompt-file argument to the holmes ask command to supply investigation prompts from a file.
    • ›Adds AWS_PROFILE support in Bedrock model requirements for named AWS profile selection.
    • ›New RabbitMQ toolset for querying RabbitMQ broker state during investigations.
    • ›New Elasticsearch toolset for querying Elasticsearch during investigations.
    • ›New Azure Kubernetes Service (AKS) toolset, including a dedicated AKS node health toolset and support for specifying the AKS environment.
    +8 moreshow less

    • ›New Git toolset enabling Holmes to inspect repository history as part of investigations.
    • ›New workload KRR (Kubernetes Resource Recommender) tool for right-sizing analysis.
    • ›Prometheus auto-discovery so Holmes can locate Prometheus instances without manual configuration.
    • ›Multi-tenant Loki and Grafana support via additional headers configuration.
    • ›Coralogix toolset gains archived log fetching and enhanced log query capabilities.
    • ›Structured output feature flag for controlling LLM response formatting.
    • ›LLM selection support, allowing users to choose which model Holmes uses.
    • ›Interactive clarifying questions and follow-up support after an investigation in the holmes ask command.
54. 0.10.9 Apr 14, 2025 · issue -367

    HolmesGPT 0.10.9 adds a by argument for investigating and enriching PagerDuty and JSM tickets.

    └──▷ GET THIS VERSION
    copy

    $ git clone --branch 0.10.9 https://github.com/HolmesGPT/holmesgpt.git
    # already have the repo? check out this version:
    $ git checkout 0.10.9

    • ›Adds by argument to the investigate and enrich commands for PagerDuty and JSM (Jira Service Management) tickets.
55. 0.10.8 Apr 10, 2025 · issue -367

    HolmesGPT 0.10.8 adds custom Prometheus label support, streaming analysis, configurable Prometheus healthchecks, and improved ArgoCD and Coralogix toolsets.

    └──▷ GET THIS VERSION
    copy

    $ git clone --branch 0.10.8 https://github.com/HolmesGPT/holmesgpt.git
    # already have the repo? check out this version:
    $ git checkout 0.10.8

    • ›Adds configurable Prometheus healthcheck support, allowing control over how Holmes validates Prometheus connectivity.
    • ›Adds support for custom labels in Prometheus queries, enabling more targeted metric lookups.
    • ›Adds streaming analysis output, delivering investigation results incrementally rather than waiting for full completion.
    • ›Improves the ArgoCD toolset with enhanced integrations for application state analysis.
    • ›Improves the Coralogix toolset to return source URLs alongside log results, making it easier to pivot directly to Coralogix.
    +1 moreshow less

    • ›Adds change history tools to support investigation workflows that incorporate recent change context.
56. 0.10.5 Mar 28, 2025 · issue -368

    HolmesGPT 0.10.5 adds Kafka, Coralogix, and Prometheus graph toolsets plus multi-cluster OpenSearch/Kafka support and streaming responses.

    └──▷ GET THIS VERSION
    copy

    $ git clone --branch 0.10.5 https://github.com/HolmesGPT/holmesgpt.git
    # already have the repo? check out this version:
    $ git checkout 0.10.5

    • ›Adds Kafka toolset (kafka) for querying Kafka clusters, with support for multiple Kafka clusters in a single configuration.
    • ›Adds initial Coralogix toolset (coralogix) for querying Coralogix logs, with prompt tuning for log retrieval.
    • ›Adds Prometheus graph-generating capability to the prometheus toolset, enabling the LLM to render metric graphs during investigations.
    • ›Adds support for multiple OpenSearch clusters in the opensearch toolset configuration.
    • ›Promotes Loki to a full-fledged log provider toolset (loki), replacing the previous limited integration.
    +9 moreshow less

    • ›Improves the Grafana Tempo toolset with enhanced trace querying capabilities.
    • ›Adds OpenSearch logs and traces support, expanding OpenSearch toolset coverage beyond search.
    • ›Adds streaming responses (ROB-717) so Holmes can stream answers to the CLI in real time.
    • ›Adds auth improvements for internet, notion, and llm toolsets.
    • ›Adds priorityClassName support to the Helm chart, allowing users to set the priority class name for Holmes pods.
    • ›Adds node selector support to the Helm chart for pod scheduling control.
    • ›Enables pod reloading when values.toolsets is updated via helm upgrade, so toolset config changes take effect without manual restarts.
    • ›Improves fetch_url tool to work with GitHub Pages sources.
    • ›Improves token counter performance via caching, reducing overhead on large investigations.
57. 0.8.1 Jan 31, 2025 · issue -370

    HolmesGPT 0.8.1 adds Gemini, ArgoCD, OpenSearch, and Grafana Loki toolsets plus AI output customization and global instructions.

    └──▷ GET THIS VERSION
    copy

    $ git clone --branch 0.8.1 https://github.com/HolmesGPT/holmesgpt.git
    # already have the repo? check out this version:
    $ git checkout 0.8.1

    • ›Adds ArgoCD toolset integration, bringing ArgoCD application state into Holmes investigations.
    • ›Adds OpenSearch toolset integration for querying OpenSearch data during investigations.
    • ›Adds Grafana Loki toolset integration for pulling logs from Loki during root-cause analysis.
    • ›Adds Gemini LLM support as a backend model option.
    • ›Adds global instructions capability, allowing operators to inject persistent context into all Holmes prompts.
    +5 moreshow less

    • ›Adds AI output customization (MAIN-2806), enabling custom sections and structured output shaping in RCA responses.
    • ›Adds kubectl with jq as a built-in tool, enabling JSON-filtered Kubernetes queries inside Holmes investigations.
    • ›Adds pod affinity configuration support to the Helm chart.
    • ›Adds optional performance logging to improve observability of Holmes internals.
    • ›Adds app diagnose chat capability for interactive application diagnosis sessions.
58. 0.7.2 Dec 10, 2024 · issue -371

    HolmesGPT 0.7.2 adds the ability to fetch deployment configuration data during investigations.

    └──▷ GET THIS VERSION
    copy

    $ git clone --branch 0.7.2 https://github.com/HolmesGPT/holmesgpt.git
    # already have the repo? check out this version:
    $ git checkout 0.7.2

    • ›Adds option to fetch deployment config, expanding the Kubernetes context available during investigations.
59. 0.7.1 Dec 6, 2024 · issue -371

    HolmesGPT 0.7.1 adds WatsonX support, kubectl events, grep log search, and memory-request inspection tools.

    └──▷ GET THIS VERSION
    copy

    $ git clone --branch 0.7.1 https://github.com/HolmesGPT/holmesgpt.git
    # already have the repo? check out this version:
    $ git checkout 0.7.1

    • ›Adds kubectl_logs_grep tool, enabling Holmes to search pod logs using grep (upgraded to grep -E extended regex) to surface relevant log lines faster.
    • ›Adds a tool to fetch top memory requests across the cluster, giving the LLM visibility into resource consumption.
    • ›Adds kubectl events capability so Holmes can retrieve and reason over Kubernetes events during investigations.
    • ›Enables the LLM to fetch logs from all containers or a specific container within a pod, expanding container-level log coverage.
    • ›Adds IBM WatsonX as a supported LLM backend, broadening model provider options.
    +2 moreshow less

    • ›Adds tag support to the ask holmes command, allowing queries to be scoped or annotated with tags.
    • ›Adds LLM evaluations framework for testing and benchmarking Holmes response quality.
60. 0.6.0 Nov 6, 2024 · issue -372

    HolmesGPT 0.6.0 adds runbook URL support, custom LLM backends, kubectl top, and a new conversation API

    └──▷ GET THIS VERSION
    copy

    $ git clone --branch 0.6.0 https://github.com/HolmesGPT/holmesgpt.git
    # already have the repo? check out this version:
    $ git checkout 0.6.0

    • ›Adds support for passing runbook URLs to the issue investigator, letting Holmes pull context from your own runbooks during investigations.
    • ›Supports custom LLM implementations, enabling teams to plug in their own LLM backends beyond the built-in providers.
    • ›Supports removing the Bedrock prefix from model name tokens for AWS Bedrock LLM configurations.
    • ›Adds a kubectl top action to the toolset, giving the AI agent access to live resource-usage metrics during investigations.
    • ›Adds a new conversation functionality, enabling multi-turn interactions with the investigator.
    +1 moreshow less

    • ›Adds an RDS single-instance lookup command to the AWS toolset for more targeted cloud investigations.
61. 0.5.0 Oct 8, 2024 · issue -373

    HolmesGPT 0.5.0 adds custom toolsets to the Helm chart and improved in-cluster authentication.

    └──▷ GET THIS VERSION
    copy

    $ git clone --branch 0.5.0 https://github.com/HolmesGPT/holmesgpt.git
    # already have the repo? check out this version:
    $ git checkout 0.5.0

    • ›Adds custom toolsets support to the Helm chart deployment.
    • ›Adds improved in-cluster authentication handling.
62. 0.4.0 Oct 3, 2024 · issue -373

    HolmesGPT 0.4.0 adds tool_calls to the /api/conversation endpoint and lets you override the UI token via environment variable.

    └──▷ GET THIS VERSION
    copy

    $ git clone --branch 0.4.0 https://github.com/HolmesGPT/holmesgpt.git
    # already have the repo? check out this version:
    $ git checkout 0.4.0

    • ›The POST /api/conversation endpoint now returns tool_calls in its response, exposing which tools Holmes invoked during an investigation.
    • ›Allows overriding the UI token with an environment variable instead of requiring it to be set in config.
    • ›Improves CLI logging and expands the --verbose flag for more detailed diagnostic output.
    • ›Improves prompt output to produce more detailed investigation results.
63. 0.3.0 Sep 2, 2024 · issue -374

    HolmesGPT 0.3.0 adds AWS/Confluence/Slab tools, a workload health endpoint, Robusta AI support, parallel tool calling, and custom CA trust.

    └──▷ GET THIS VERSION
    copy

    $ git clone --branch 0.3.0 https://github.com/HolmesGPT/holmesgpt.git
    # already have the repo? check out this version:
    $ git checkout 0.3.0

    └──▷ USE IT
    Set the LLM model once in config rather than passing it on every CLI invocation.
    yaml copy

    model: gpt-4o

    • ›Adds model parameter support in the config file, allowing the LLM model to be set without a CLI flag.
    • ›Adds AWS, Confluence, internet, and Slab tools, expanding the set of data sources Holmes can query during investigations.
    • ›Adds a workload health API endpoint with JSON response, enabling programmatic health checks via the Holmes server.
    • ›Adds a chat input box API for programmatic interaction with Holmes.
    • ›Adds support for Robusta AI as an LLM backend for Holmes.
    +5 moreshow less

    • ›Adds support for trusting custom CA certificates, enabling use with private TLS infrastructure.
    • ›Adds tolerations support for the Holmes pod, allowing scheduling on tainted Kubernetes nodes.
    • ›Enables parallel tool calling, allowing Holmes to invoke multiple tools simultaneously for faster investigation.
    • ›Adds K9s plugin integration, letting users invoke Holmes directly from the K9s terminal UI.
    • ›Increases accuracy of holmes ask results through improved prompting.
64. 0.2.6 Jul 24, 2024 · issue -376

    HolmesGPT 0.2.6 adds Slack output for the ask command, streamlined Amazon Bedrock support, and LiteLLM env var support in Helm.

    └──▷ GET THIS VERSION
    copy

    $ git clone --branch 0.2.6 https://github.com/HolmesGPT/holmesgpt.git
    # already have the repo? check out this version:
    $ git checkout 0.2.6

    • ›Helm chart now supports LiteLLM environment variables for configuring LLM backends.
    • ›Adds Slack as a destination for the ask command, enabling query results to be sent directly to Slack channels.
    • ›Streamlines Amazon Bedrock integration alongside other LLM models for a more seamless setup experience.
    • ›Adds a post-processing option for verbose ('chatty') AI responses.
    • ›Updates PagerDuty integration with minor improvements.
    +2 moreshow less

    • ›Adds Prometheus Operator RBAC rules for Holmes to the service account template in the Helm chart.
    • ›Improves the Kubernetes toolset with minor enhancements.
65. 0.2.5 Jun 24, 2024 · issue -377

    HolmesGPT 0.2.5 adds a --file flag and ships as a pip-installable Python package.

    └──▷ GET THIS VERSION
    copy

    $ git clone --branch 0.2.5 https://github.com/HolmesGPT/holmesgpt.git
    # already have the repo? check out this version:
    $ git checkout 0.2.5

    └──▷ TRY IT
    Feed a pre-written investigation prompt from a file instead of typing it inline on the command line.
    copy

    $ holmes ask --file <path>

    • ›Adds --file flag to pass input from a file directly to the CLI.
    • ›Makes HolmesGPT installable as a Python package via pip.
66. 0.2.4 Jun 17, 2024 · issue -377

    HolmesGPT 0.2.4 adds label-based filtering for Alertmanager alert ingestion.

    └──▷ GET THIS VERSION
    copy

    $ git clone --branch 0.2.4 https://github.com/HolmesGPT/holmesgpt.git
    # already have the repo? check out this version:
    $ git checkout 0.2.4

    • ›Adds label-based filtering for Alertmanager, allowing users to scope alert ingestion to specific label sets.
67. 0.2.2 Jun 14, 2024 · issue -377

    HolmesGPT 0.2.2 adds OpsGenie and PagerDuty investigation commands and JSON output flag.

    └──▷ GET THIS VERSION
    copy

    $ git clone --branch 0.2.2 https://github.com/HolmesGPT/holmesgpt.git
    # already have the repo? check out this version:
    $ git checkout 0.2.2

    • ›Adds holmes investigate opsgenie command to investigate OpsGenie alerts directly from the CLI.
    • ›Adds holmes investigate pagerduty command to investigate PagerDuty incidents directly from the CLI.
    • ›Adds a CLI flag to write investigation output to JSON for machine-readable results.