Skip to content
Get packs

Guide · Agent skill

How to use the Org Health Diagnostic agent skill

Updated · written with AI from the project's own documentation and checked against our catalogue data

The Org Health Diagnostic is an agent skill that teaches an AI agent to run a company-wide health check. You supply your key metrics, and the agent scores eight areas (financial, revenue, product, engineering, people, operations, security and market) on a traffic-light scale with drill-down recommendations. This guide covers what the skill does, how to install it, how to trigger it, and what to watch for before relying on the results.

Org Health Diagnosticalirezarezvani/claude-skills · MIT

Quick start

  1. 1Download the ZIP. Get the org-health-diagnostic ZIP from the skill's page on this directory.
  2. 2Unzip the skill. Place the folder in ~/.claude/skills/ for personal use or in the project's .claude/skills/ folder.
  3. 3Or upload it in the apps. In the Claude apps, go to Settings → Capabilities → Skills and upload the ZIP.
  4. 4Check Python 3. Confirm Python 3 is installed; the script needs no extra packages.
  5. 5Gather your metrics. Collect the figures you have, such as runway, NRR, NPS, deployment frequency and attrition. Missing ones are allowed.
  6. 6Ask for a health check. Say something like 'Run an org health check' and paste your metrics, or run the script directly.
  7. 7Review priorities. Read the traffic lights, top priorities and watch list, then verify the inputs before acting.

What is an agent skill and what does this one do?

An agent skill is a folder of instructions and optional scripts that an AI agent loads when a task matches the skill's description. This one is designed for company-wide reviews rather than a single metric.

You provide the numbers. The agent then scores each of eight dimensions from 1 to 10 and labels it green (7-10), yellow (4-6) or red (1-3). It produces a weighted overall score based on company stage and points out which related dimensions are likely to deteriorate next.

  • Financial (CFO): runway, burn multiple, gross margin, MoM growth, revenue concentration
  • Revenue (CRO): net revenue retention, logo churn, pipeline coverage, CAC payback, ACV trend
  • Product (CPO): NPS, DAU/MAU ratio, core feature adoption, time-to-value, CSAT
  • Engineering (CTO): deployment frequency, change failure rate, MTTR, tech debt ratio, incident frequency
  • People (CHRO): regrettable attrition, engagement score, time-to-fill, manager-to-IC ratio, internal promotion rate
  • Operations (COO): OKR completion, decision cycle time, meeting effectiveness, process maturity, cross-functional initiative completion
  • Security (CISO): security incidents, compliance status, vulnerability remediation SLA, training completion, pen test recency
  • Market (CMO): CAC trend, organic vs paid lead mix, win rate, competitive win rate, brand NPS

How do you install the Org Health Diagnostic skill?

Download the ZIP from the skill's page on this directory and unzip it. For a personal setup, place the folder in ~/.claude/skills/. For a single project, place it in that project's .claude/skills/ folder.

You can also upload the ZIP in the Claude apps under Settings → Capabilities → Skills.

  • Personal install: unzip into ~/.claude/skills/
  • Project install: unzip into the project's .claude/skills/
  • Claude apps: Settings → Capabilities → Skills, then upload the ZIP

What do you need before you start?

The skill needs Python 3. The included script uses only the Python standard library, so there are no packages to install. You also need your own company metrics; the skill does not connect to external accounts or require API keys, and the source does not mention any network use.

The skill's files are SKILL.md, scripts/health_scorer.py and references/health-benchmarks.md.

  • Python 3 installed on your machine
  • Your company metrics, entered by you
  • No API keys or external accounts mentioned in the source

How do you trigger the skill with natural requests?

You do not need special syntax. Ask in plain language and include your metrics. The skill's description names these triggers: assessing overall company health, preparing for board reviews, identifying at-risk functions, and requests that mention org health, a health check or a health dashboard.

For a single area, name the dimension in your request. The skill lists the dimension names as options you can use in a request.

  • "Run an org health check" and paste your key metrics
  • "Prepare an org health dashboard for our board review"
  • "Score our people health" for a single-dimension deep dive
  • "Identify which functions are at risk this quarter"

How do you run the guided command-line version?

The script can be run directly for a guided session where you enter metrics and receive a scored dashboard. Add the --json flag if you want raw output for another tool to read.

python scripts/health_scorer.py python scripts/health_scorer.py --json

What does the dashboard output look like?

The dashboard shows an overall score with a trend arrow, one line per dimension with its score and key figures, then a short list of top priorities. Each priority has an owner and a suggested action, and a watch list shows likely cascade risks.

Scores depend on the thresholds in the benchmark file. Stage matters: the benchmarks are set separately for Seed, Series A, Series B and Series C, and the overall score is weighted by stage.

  • Dimension scores with traffic lights
  • Top priorities with suggested owners and actions
  • Watch list showing which dimensions may deteriorate next

What about missing data and safety?

You do not need every metric. Missing metrics are excluded from the score and flagged as data needed, so the score still runs on what you provide.

The skill runs scripts, so read health_scorer.py before running it on sensitive numbers. The source says the script uses only the standard library and that the skill does not use the network. The security section only checks self-reported status, so it is not a security audit or compliance certification.

  • Runs scripts: yes, the Python script in scripts/
  • Uses network: no, according to the source
  • Security scores reflect self-reported status only

What people use it for

  • Preparing a quarterly snapshot to compare company health over time
  • Getting ready for a board review with a consistent scorecard
  • Spotting which function is slipping before it becomes a crisis
  • Running a single-dimension deep dive on people or engineering

Frequently asked questions

Is the Org Health Diagnostic skill free?

The skill is released under the MIT licence, as stated in its source. The source does not mention pricing for any other service, so check the directory page for current terms.

What do I need to run the Org Health Diagnostic skill?

You need Python 3, because the script uses only the standard library. You also need your own company metrics, since the skill does not connect to external accounts or require API keys.

How do I install an agent skill like this one?

Download the ZIP from the skill's page and unzip it into ~/.claude/skills/ for personal use or into a project's .claude/skills/ folder. You can also upload the ZIP in the Claude apps under Settings → Capabilities → Skills.

Does the skill send my company data anywhere?

The source states that the skill does not use the network. Scripts do run locally, so review health_scorer.py if your numbers are sensitive.

Can I use it if my company is not SaaS?

The source warns against this. Many thresholds such as NRR, ACV and MRR-style growth assume subscription software, so results may be less meaningful for other business models.

Can I trust the scores for legal or compliance decisions?

No. The source says the security section only checks self-reported status, and the benchmark thresholds are generic rules of thumb. Verify financial and HR data independently before making decisions.

New agent skills, repos and AI tools — every week

The best new skills and open-source projects, the fastest-growing repos, and one free prompt. Five minutes to read. Unsubscribe anytime.

Our verdict

This skill suits founders and executives preparing board reviews or quarterly snapshots, and leaders who want a consistent template to track health over time. It is not a substitute for verified financial, HR or compliance data, and it is a poor fit for companies outside the SaaS model.

Download the skill