LoginSign Up
← Back to all sparks
H

Honeycomb

INFRA · APIS
Velocity7.5

Canvas agents gain memory, and onboarding moves into the editor

observabilitycanvas-agentsanomaly-detectionmcpopentelemetryauto-investigation
◆Current state
Honeycomb is building an investigation agent rather than a query tool. Automatic Investigations now dispatch to Canvas agents carrying awareness of recent firings of the same Trigger, Burn Alert or Anomaly, so repeat issues get pinpointed against prior hypotheses. Around that sit Anomaly Detection in open beta, MCP-based onboarding that instruments a codebase from the editor, Canvas connectors for Linear and GitHub, and telemetry stats in the Activity Log.
◆Where it's heading
Every recent release reduces what a human has to know before Honeycomb is useful. Detection needs no thresholds, onboarding needs no manual SDK setup, and now the agent retains context across alert firings instead of starting cold each time. Canvas is becoming the product's centre of gravity — the surface that reads connectors, edits Triggers and SLOs, and accumulates conclusions.
◆Prediction
Anomaly Detection should widen beyond error rate and presence to latency and request rate as it approaches GA, and the alert-history awareness added here is the groundwork for agents that correlate across different alerts rather than repeat firings of one.

◆Recent moves

  1. 27d ago

    Response Awareness in Automatic Investigations

    Automatic Investigations now reach Canvas agents with awareness of recent firings of the same alert, letting them build on earlier hypotheses instead of re-deriving them. It is the memory layer under the auto-investigation arc — incremental on its own, but it is what makes a repeat incident cheaper than the first.

    View source ↗
  2. 1mo ago

    New: Onboard with Honeycomb MCP

    Onboarding moves to the Honeycomb MCP server: connect a codebase from the editor, auto-instrument through a guided prompt, and send live or synthetic data without manual SDK setup. This attacks the setup cost that Anomaly Detection's zero-configuration promise depends on.

    View source ↗
  3. 1mo ago

    Anomaly Detection: Now in Beta

    ⚡ SPARK

    Anomaly Detection opened in beta for every team, baselining each service from existing OpenTelemetry data with nothing to configure. It anchors the current arc: the other releases in this window exist to feed it data and to act on what it finds.

    View source ↗
  4. 1mo ago

    Activity Log now includes telemetry stats

    The Activity Log environment gains a telemetry_stats dataset recording event volume, acceptance outcomes and errors per environment, dataset and API key, with a board template to query it. Honeycomb turning its own ingest pipeline into queryable data, useful for the data-quality debugging that self-serve onboarding will generate.

    View source ↗
  5. 1mo ago

    Edit Triggers, SLOs, and Boards in Canvas

    Canvas can now create and update Triggers, SLOs and Boards, with human approval required and the capability withheld during automatic investigations. The agent gains write access to configuration under an explicit approval gate — a deliberate line between what it may propose and what it may do unattended.

    View source ↗
  6. 1mo ago

    Honeycomb Canvas Connectors: Now in Beta

    ⚡ SPARK

    Connectors opened in beta with Linear and GitHub, letting Canvas read outside tools during an investigation. This is the release that turned Canvas from a Honeycomb feature into a workspace, and the alert-memory and edit capabilities since have built on that footing.

    View source ↗