mcpField
S
score

score

github.com/hainrixzData & databases2270 installsUpdated 6d ago
Repository

About score

score is a skill for scoring database audit findings. It reruns `scripts/score.mjs` to produce two separate scores and their breakdowns, using either a findings file or the latest audit from the session.

Key features

  1. 01Recompute two audit scores
  2. 02Accept findings JSON files
  3. 03Select a database paradigm profile
  4. 04Show category values and weights
  5. 05Report severity caps and confidence

Use cases

  1. 01Review scores from the latest database audit
  2. 02Rescore findings with another paradigm profile
  3. 03Inspect category scores and severity caps
  4. 04Check the count of findings needing API verification

Install

Copy the command and run it in a terminal.

$npx skills add https://github.com/hainrixz/claude-db --skill score

After installing, type/scorein Claude Code to confirm it loaded.

Details

From the project README

/claude-db:score

Recompute and show the two 0–100 scores from a set of findings by re-running scripts/score.mjs (pure, fully reproducible — it implements references/scoring-model.md).

  • If a findings JSON path is given in $ARGUMENTS, score that file: node scripts/score.mjs --findings <path> [--paradigm <p>]. The file is a JSON array of findings (or { "findings": [...] }) conforming to schema/finding.schema.json.
  • Otherwise use the findings from the most recent audit this session. If no audit has run yet, tell the user to run /claude-db:audit <target> first — do not invent findings.
  • The --paradigm flag selects the category profile (PROFILES in score.mjs); the default is relational. Re-scoring under a different profile is how you show, e.g., the same Mongo findings under the document profile (no foreign-key categories in the denominator).

Show both scores with bands (A ≥ 90 · B ≥ 80 · C ≥ 70 · D ≥ 60 · F < 60), the per-category breakdown (value × weight, active?), any severity-gating cap (capped:true with the uncapped computed shown), and the needs_api count as score confidence. needs_api and speculative findings never cap. Two scores, never blended.

Comments

Markdown supported