Score Prismstack product skills against a 15D rubric and save results.
npx skills add https://github.com/fagemx/prismstack --skill self-check-fagemx
Please help me install this Agent Skill. Skill: self-check Source: https://github.com/fagemx/prismstack/tree/main/.claude/skills/self-check Command: npx skills add https://github.com/fagemx/prismstack --skill self-check-fagemx
This internal skill audits Prismstack's own product skills to ensure quality and consistency with a shared 15D rubric.
Run the self-check skill to audit Prismstack's skills and generate the latest results.
High-intent search queries and answers about installing and using this skill.
To audit skill quality using a rubric, this skill identifies and scores product skills against a published 15D rubric, analyzing gaps to ensure consistency. It automatically discovers skills via the SKILL.md manifest and evaluates them across all 15 dimensions.
Yes, you can track quality audit changes over time by comparing the current run against the last run. The skill calculates a delta since the previous audit and saves the consolidated results to a local last-results.json file for tracking.
A 15D rubric for skill quality assessment is a multidimensional scoring framework used to evaluate product skills. It scores each skill across 15 distinct dimensions, aggregates the results, and generates a cross-skill pattern analysis to identify quality gaps.
You generate a consolidated report for internal audits by running the skill to evaluate all discovered SKILL.md files. It produces a Phase 3 score table with per-skill scores and a delta since the last run, outputting the final report to last-results.json.
No, you do not need external dependencies to perform a self-check audit. The skill operates independently by loading the quality rubric directly from the repository and discovering product skills via the standard manifest without requiring additional components.
A limitation of automated skill metrics analysis is that it is restricted to evaluating skills with a standard SKILL.md manifest. It relies entirely on the predefined 15D rubric loaded from the repository, meaning custom dimensions or non-standard formats are not supported.
swarm-ai-researchGenerate four skill variations, score them, and commit the winner as a PR.
Obsidian-OwlDetect orphaned exports and wiring gaps in frontend modules.
kunalrawat425Aggregate type checks, linting, tests, and dead-code detection into a weighted 0-10 health score.
thktAudit SOW/Spec documents against a 100-point scoring framework.
