SuperAPI AI Directory

cli-analysis

Coding & development

View repository

Updated 10-16 16:26

About this skill

Code Quality Analysis with the polyscan CLI

Use the polyscan command-line tool when the task needs report files, CI integration, or machine-readable results. It analyzes JavaScript, TypeScript, Go, Rust and C++ in one run; the language of each file is detected from its extension.

No install needed: npx polyscan@latest analyze <path> (or npm install -g polyscan for a permanent install). The examples below use a bare polyscan for readability — prefix with npx polyscan@latest when polyscan isn't installed.

analyze — the one command

polyscan analyze .                            # HTML report (polyscan-report.html) + opens browser
polyscan analyze --no-open .                  # HTML report without opening a browser
polyscan analyze --format json src/           # JSON results to stdout
polyscan analyze --format text src/           # human-readable text to stdout
polyscan analyze --select complexity,clone .  # only specific analyses
polyscan analyze --no-open -o report.html .   # custom report path
polyscan analyze --min-complexity 10 .        # list only functions at or above 10

Key flags:

  • --format html|json|text (-f): output format; json and text write to stdout instead of an HTML report
  • --select (-s): complexity,deadcode,clone,cbo,lcom,deps (all run by default); deps applies to Go and JavaScript/TypeScript; cbo to Go, Rust and JavaScript/TypeScript; lcom to Go and Rust; deadcode to JavaScript/TypeScript only
  • --no-open: don't auto-open the HTML report in a browser (use in scripts)
  • -o <path>: HTML report path (default: polyscan-report.html in the current directory)
  • --min-complexity <n>: hide functions below this complexity from the listing (scores still cover everything)
  • --exclude <patterns>: leave files and directories out of every analysis; a glob without a slash matches a file name or a directory anywhere on the path (fixtures), one with a slash matches a path relative to the analyzed directory with ** for any number of segments (src/generated/**)
  • --include-tests: analyze test files and test code, which are left out by default (*_test.go, *.test.*, *.spec.*, __tests__/, Rust #[cfg(test)] and tests/, C++ *_test.* and test/)

With --format json, the machine-readable results go to stdout and the health-score summary goes to stderr, so stdout stays parseable.

Language coverage

Complexity and clone detection cover every supported language. Dependency analysis covers Go and JavaScript/TypeScript. Class coupling (CBO) covers Go, Rust and JavaScript/TypeScript, and class cohesion (LCOM4) Go and Rust. Dead code exists for JavaScript/TypeScript only. The health score is computed over the dimensions that ran: a dimension a language does not have is left out, not scored as clean.

CI usage

There is no separate gate command; gate on the JSON output:

polyscan analyze --format json src/ > report.json
jq -e '.summary.health_score >= 80' report.json

Configuration

JavaScript/TypeScript analysis reads jscan.config.json (or .jscanrc.json) from the analyzed project when present — complexity thresholds and exclude patterns carried over from jscan, which merged into polyscan. There is no config file for the other languages yet.

Reporting Results

Summarize the health score and grade, list the specific functions/files behind each failing category, and suggest fixes. For CI setup, recommend a JSON gate as above plus periodic polyscan analyze HTML reports for deep reviews.