skills.rest
burningportra
burningportraCOMMUNITY
Explore all skills by this author →
GitHub

flywheel-healthcheck

Audits flywheel dependencies, codebase state, and bead graph into a scored health report.

3|1|Updated Apr 15, 2026
Search Tags:#claude code#multi-agent#health check#diagnostics#dependency check#codebase audit#bead tracker
One-click install
npx skills add https://github.com/burningportra/agent-flywheel-plugin --skill flywheel-healthcheck-burningportra
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: flywheel-healthcheck
Source: https://github.com/burningportra/agent-flywheel-plugin/tree/main/skills/flywheel-healthcheck
Command: npx skills add https://github.com/burningportra/agent-flywheel-plugin --skill flywheel-healthcheck-burningportra

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Multi-agent coding workflows depend on a chain of external tools (bead tracker, bead visualizer, agent-mail server, MCP server) and on-disk state that can silently drift or break. Without a periodic audit, developers discover broken dependencies, orphaned beads, or accumulating technical debt only when the workflow stalls mid-task. ## Core Features & Use Cases - Dependency verification: Checks versions of br and bv, pings the agent-mail liveness endpoint, inspects git status, checkpoint state, and MCP server build output in parallel. - Codebase health scan: Counts TODO/FIXME/HACK markers, computes test-to-source file ratios, and surfaces lint or compilation issues via an Explore agent. - Bead graph audit: Runs br list and bv to count open/closed/in-progress/deferred beads and detect cycles or orphaned beads, then renders a formatted report with a health score out of 10. - Use Case: Run it on a weekly cadence to confirm the agent-flywheel toolchain is healthy before starting a large multi-agent implementation wave. ## Quick Start Ask the agent to run a full flywheel health check and report the health score for dependencies, codebase, and beads.

Frequently Asked Questions about flywheel-healthcheck

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I check the health of an agent-flywheel setup?▼

Run the flywheel-healthcheck skill, which verifies br and bv versions, pings the agent-mail liveness endpoint, checks git status and checkpoint state, and audits the bead graph. It outputs a formatted report with a health score out of 10.

What does the flywheel health check report include?▼

The report covers three sections: dependencies (br, bv, agent-mail, MCP server build), codebase metrics (TODO count, test-to-source ratio, git cleanliness), and bead status (open/closed counts, cycles, orphans), ending with an overall score out of 10.

When should I run healthcheck instead of flywheel-doctor?▼

Run flywheel-doctor first for a read-only snapshot when you suspect toolchain drift, then flywheel-setup to apply fixes. Use healthcheck as a periodic deep audit of codebase, bead graph, and dependencies, not for fresh-clone setup problems.

Why does the health check report the MCP server as not built?▼

The check looks for mcp-server/dist/server.js and warns when it is missing. Fix it by running npm run build inside the mcp-server directory, then rerun the health check to confirm.

Can the health check detect problems in the bead dependency graph?▼

Yes. It runs br list --json to count beads by status and bv --json to detect cycles and orphaned beads, flagging them in the BEADS section of the report.

Related Skills

Explore More Software Engineering Skills→
DicklesworthstoneDicklesworthstone
70

agent-swarm-workflow

Coordinate parallel software tasks across multiple agents using Agent Mail and BV prompts.

Community
Advanced
JDHayesBCJDHayesBC
1

canopy

Survey codebase health, active work, and debt before changes.

Community
Intermediate
pertain99pertain99
1

healthcheck

Audit host security and configure risk tolerance for OpenClaw deployments.

Community
Advanced
tiankong0101-bytetiankong0101-byte

autonomous-loop

Monitor long-running automated processes and recover them from stalls or deadlocks.

Community
Advanced