oss-project-health

Assess open-source repository maintenance vitality using CHAOSS metrics and OpenSSF Scorecard checks.

4|1|Updated Jul 30, 2026
One-click install
npx skills add https://github.com/radarist/structured-analytic-skills --skill oss-project-health-radarist
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: oss-project-health
Source: https://github.com/radarist/structured-analytic-skills/tree/main/skills/oss-project-health
Command: npx skills add https://github.com/radarist/structured-analytic-skills --skill oss-project-health-radarist

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) components.

What problem does it solve? Teams often adopt open-source dependencies based on star counts, only to discover later that a project is maintained by one burned-out contributor or carries unpatched security advisories. This Skill reads a repository's actual maintenance vitality — contributor concentration, commit and release cadence, issue responsiveness, and open advisories — and returns a defensible healthy / at-risk / abandoned verdict. ## Core Features & Use Cases - Bus factor analysis: Computes the CHAOSS Contributor Absence Factor (bots excluded) and the Elephant Factor over organisations to expose single-maintainer or single-company dependence. - Cadence and responsiveness scoring: Rates days since last commit, release frequency, median issue first-response time, PR merge time, and issue backlog ratio against documented green/amber/red thresholds. - Security and hygiene checks: Flags advisories unpatched beyond a 90-day fix window, missing licences, and absent SECURITY.md, CI, or code of conduct. - Deterministic companion tool: scripts/osshealth.py (stdlib only) computes the full scorecard, prints the report template, and can fetch metrics from the GitHub REST API. - Use Case: Before betting production infrastructure on a vector database, run the vitality read on both candidate repositories and compare their verdicts and biggest risks side by side. ## Quick Start Ask the assistant to run an OSS project health check on a specific owner/repo slug, covering bus factor, commit and release cadence, issue response time, and open security advisories, and to return a healthy, at-risk, or abandoned verdict.

Frequently Asked Questions about oss-project-health

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

FAQPage Schema
How do I check if a GitHub repository is still maintained?▼

Gather a 12-month snapshot of commits by author, days since last commit, releases, median issue first-response time, and open advisories, then rate each against documented thresholds. The verdict is healthy, at-risk, or abandoned based on red and amber indicator counts, never on star count.

How is bus factor calculated for an open-source project?▼

Bus factor is the CHAOSS Contributor Absence Factor: rank authors by commits in the window, exclude bot accounts, and count the smallest number of authors whose commits reach 50 percent of the total. A factor of 1 is fragile, 2 is amber, and 3 or more is spread.

What is the difference between project health and technology readiness?▼

Project health measures current maintenance capacity — who is still doing the work — while TRL measures deployment maturity. A TRL-9 project can still be one resignation from unmaintained, so the two assessments are orthogonal and complementary.

Does a low commit count mean a project is abandoned?▼

No. A feature-complete library commits rarely and still lives. Abandoned requires no commits for over 365 days, no release for over 365 days, and median issue first response over 14 days, all three known; check releases and responsiveness before concluding.

Can I run the health assessment without network access?▼

Yes. The companion script osshealth.py is standard-library only and deterministic; the assess subcommand works offline from a metrics JSON file. Only the fetch subcommand touches the network, calling the GitHub REST API for an explicit owner/repo.

When should I not use repository health metrics?▼

Do not use them for market narrative questions like whether a technology is overhyped, which belongs to hype-cycle analysis, or for single release-note parsing. Also avoid forcing a verdict when the project has no public repository or most data fields are null.