claude

cardinal

Connect Claude Code to Cardinal: telemetry to the Outcomes Dashboard. Unified Cardinal MCP coming in a future release.

0 on GitHubby CardinalHQv0.63.4Apache-2.0updated todayobservability
/

Plugin cardinal from marketplace cardinalhq-claude-plugin in cardinalhq/cardinal-claude-plugin. Commands read from the repo's manifest — today.

This plugin runs code on your machine — its hooks fire automatically on events like tool use and it starts local MCP servers. The exact commands are listed below; read them before installing.

What it installs

Skills (11)

  • checks

    Build and run a Cardinal check (preview) — a reusable, evidence-backed decision made for every trace in a time range, with a contract the user approves, approved examples, and per-trace results with evidence. Use when answering needs a judgment about each trace that one aggregate query cannot give —

  • canvas

    Draw and preview the Canvas of a Cardinal Investigation Storyboard scene — free-form HTML/CSS/SVG/d3/Plot in Cardinal's sandboxed frame, with the five cv operations (cv.data, cv.mark, cv.embed, cv.reveal, cv.highlight) and the in-frame prefabs (timeline, trace-waterfall, world-graph, compare, diff)

  • connect

    Connect Claude Code to Cardinal — runs the device-code flow to enable telemetry to the Outcomes Dashboard AND the unified Cardinal MCP tools, in one consent.

  • decision

    Check whether Cardinal decision capture is on, or turn it on or off, with the plugin's cardinal-decision command. Use only when the user asks about Cardinal decision capture (recording the agent's decisions to Cardinal), not for making or recording a decision.

  • disconnect

    Disconnect this Claude Code install from Cardinal — revoke the ingest, MCP and control-plane keys, strip the plugin's env block, delete local state and caches.

  • install-site

    Install a Cardinal site (perch + optional POC Lakerunner) into a Kubernetes cluster you have access to, driven by the Cardinal control-plane API.

  • migrate-from-grafana

    Migrate Grafana dashboards and alert rules into Cardinal (Maestro dashboards + lakerunner alert rules). Use whenever someone wants to move, copy, import, port or switch their Grafana (Cloud, OSS or Enterprise) dashboards or alerts to Cardinal / CardinalHQ / lakerunner / Maestro, bring existing Grafa

  • optimize-toolkit

    Mine this engineer's own session telemetry for capability-fit recommendations across agents, skills, commands, and MCP tools (extract a new sub-agent or command, pin/downgrade a model tier, adopt or swap an existing capability, consolidate duplicates) and, on confirmation, write the accepted artifac

  • status

    Verify the Cardinal plugin's wiring on this Claude Code install (telemetry and/or MCP).

  • storyboard

    Publish a Cardinal storyboard by updating its evidence-bound scenes and visualization, rendering and reviewing previews, then changing the draft status to published. Also use for explicitly requested private visualization updates. Milestones are written by this session on its checkpoints; Cardinal n

  • supervise

    Supervise another agent session through its Cardinal Investigation — watch everything it does (its checkpoints and its captured tool activity), validate every finding, question every hypothesis, challenge what is wrong or unreported, and review its artifacts when it is done, posting verdicts the own

Subagents (2)

  • brainstorm

    Self-contained divergent idea generation — delegate when you need N candidate approaches, designs, names, or solutions with tradeoffs and a recommendation, and the request is fully specifiable up front (fire-and-forget). Give it the problem, hard constraints, and how many options you want; it return

  • stepback

    Only for the Cardinal supervise skill's fresh-context passes. Reads a brief and the code; changes nothing.

MCP servers (2)

  • cardinal

    connects to: ${CARDINAL_MCP_URL}

  • cardinal-behavior

    runs: python3 ${CLAUDE_PLUGIN_ROOT}/bin/cardinal-behavior

Hooks (29)

  • UserPromptSubmit

    ${CLAUDE_PLUGIN_ROOT}/hooks/git-state.py

  • UserPromptSubmit

    ${CLAUDE_PLUGIN_ROOT}/hooks/limits-gate.py

  • UserPromptSubmit

    ${CLAUDE_PLUGIN_ROOT}/hooks/owner-input.py

  • UserPromptSubmit

    ${CLAUDE_PLUGIN_ROOT}/hooks/owner-relay.py

  • UserPromptSubmit

    ${CLAUDE_PLUGIN_ROOT}/hooks/decision-prompt.py

  • UserPromptSubmit

    ${CLAUDE_PLUGIN_ROOT}/hooks/storyboard-discovery.py

  • PreToolUse

    ${CLAUDE_PLUGIN_ROOT}/hooks/invariant-check.py

  • PreToolUse

    ${CLAUDE_PLUGIN_ROOT}/hooks/storyboard-edit-lookup.py

  • PreToolUse

    ${CLAUDE_PLUGIN_ROOT}/hooks/storyboard-edit-lookup.py

  • PreToolUse

    ${CLAUDE_PLUGIN_ROOT}/hooks/storyboard-context.py

  • PreToolUse

    ${CLAUDE_PLUGIN_ROOT}/hooks/checks-gate.py

  • PostToolUse

    ${CLAUDE_PLUGIN_ROOT}/hooks/subagent-usage.py

  • PostToolUse

    ${CLAUDE_PLUGIN_ROOT}/hooks/storyboard-preview.py

  • PostToolUse

    ${CLAUDE_PLUGIN_ROOT}/hooks/storyboard-hero.py

  • PostToolUse

    ${CLAUDE_PLUGIN_ROOT}/hooks/storyboard-token.py

  • PostToolUse

    ${CLAUDE_PLUGIN_ROOT}/hooks/evidence-capture.py

  • PostToolUse

    ${CLAUDE_PLUGIN_ROOT}/hooks/investigation-events.sh

  • PostToolUseFailure

    ${CLAUDE_PLUGIN_ROOT}/hooks/evidence-capture.py

  • PostToolUseFailure

    ${CLAUDE_PLUGIN_ROOT}/hooks/investigation-events.sh

  • SessionStart

    ${CLAUDE_PLUGIN_ROOT}/hooks/initiative-convention.py

  • SessionStart

    ${CLAUDE_PLUGIN_ROOT}/hooks/checks-gate.py

  • SessionStart

    ${CLAUDE_PLUGIN_ROOT}/hooks/storyboard-session.py

  • SessionStart

    ${CLAUDE_PLUGIN_ROOT}/hooks/storyboard-discovery.py

  • SessionStart

    ${CLAUDE_PLUGIN_ROOT}/hooks/plan-state.py

  • SubagentStart

    ${CLAUDE_PLUGIN_ROOT}/hooks/storyboard-discovery.py

  • Stop

    ${CLAUDE_PLUGIN_ROOT}/hooks/turn-usage.py

  • Stop

    ${CLAUDE_PLUGIN_ROOT}/hooks/plan-usage.py

  • Stop

    ${CLAUDE_PLUGIN_ROOT}/hooks/investigation-events.sh

  • Stop

    ${CLAUDE_PLUGIN_ROOT}/hooks/owner-relay.py stop

View source on GitHub cardinalhq.io

Similar plugins

CMEM Pro

claude

CMEM Pro for the Claude app - hooks stream tool use to cmem.ai and inject observations into new sessions and agents

3 skills · 1 MCP server · hooks

thedotmack/claude-mem 116,618

Structured logging, distributed tracing, and metrics — correlate agent swarm activity with application telemetry

2 skills · 1 command · 1 agent

ruvnet/ruflo 81,589

honeycomb

claudecodex

Skills, agents, and workflows for Honeycomb observability — query patterns, production investigations, SLOs, OpenTelemetry instrumentation, and Beeline migration. Designed to complement the Honeycomb MCP server.

11 skills · 1 command · 2 agents · 1 MCP server · hooks

anthropics/claude-plugins-official 47,076

Kubernetes manifest generation, networking configuration, security policies, observability setup, GitOps workflows, and auto-scaling

4 skills · 1 agent

wshobson/agents 42,461

Honeycomb

claudecodex

Skills, agents, and workflows for Honeycomb observability — query patterns, production investigations, SLOs, OpenTelemetry instrumentation, and Beeline migration. Designed to complement the Honeycomb MCP server.

11 skills · 1 command · 2 agents · 1 MCP server · hooks

anthropics/knowledge-work-plugins 35,601

Maintain cardinal?

This page updates itself from your repo. Add the badge to your README, or get it Featured to pin it above the organic results with a followed link to your site.

claude & codex pluginslisted