tech-stack-recommender

I've been poking around tech stack selection problems lately, and I see you're tackling the decision-making side of it—how did you land on this particular approach for guiding those choices?

Links:

• GitHub
• Your agentic skill in the SkillzWave marketplace

The TL;DR

You're at 70/100, solidly in C territory. This is based on Anthropic's skill grading rubric across five pillars. Your strongest area is Spec Compliance (12/15)—the frontmatter and naming are clean. Weakest area is Progressive Disclosure Architecture (15/30)—you're packing 448 lines into a single file when you could be more strategic about layering content.

What's Working Well

• Clear decision framework (lines 20-41) with ASCII diagrams showing inputs and outputs—this actually gives people a visual anchor for how to use the skill
• Excellent stack templates (lines 187-307) that show real-world examples for SaaS, e-commerce, and enterprise scenarios—people can see themselves in these
• Strong metadata and triggers in the frontmatter—terms like "selecting frameworks," "language selection," and "infrastructure choices" are solid discovery hooks
• Practical evaluation checklist (lines 327-348) that gives people a way to validate their choice after the fact

The Big One: Missing Reference Files

You're listing two reference files that don't actually exist: framework-comparison.md and migration-playbooks.md (lines 444-447). This is a PDA violation because you're claiming a layered structure you haven't actually built.

Here's the fix: Either create those files and move your detailed comparison content into them, or remove the References section. If you go with creating them, you'd move those 4 stack templates (120 lines of repetition) into references/stack-templates.md and link to it. That alone would free up space and fix your token economy. Potential gain: +5 points.

Other Things Worth Fixing

1. Template repetition is eating your token budget (lines 185-307)—four nearly identical 30-line templates with the same structure (Frontend, Backend, Database, DevOps). Keep one in-SKILL.md example, move the rest to a reference file. +3 points.

2. No step-by-step workflow exists—you show a decision framework diagram but not numbered steps for how someone actually uses it. Add something like: "1. Gather requirements → 2. Assess team factors → 3. Match to Quick Stack table → 4. Validate trade-offs." +2 points.

3. Mixed voice and tense throughout—line 8 says "Provides structured recommendations" (declarative), but lines 330-332 use questions ("Current skills alignment?"). Pick one voice and stick with it. +1 point.

4. No feedback loop for validation—the checklist exists, but there's no guidance on what to do if the recommendation isn't working. Add a simple section: "If adoption is slow, assess learning curve; if performance lags, revisit database choice." +2 points.

Quick Wins

• Create framework-comparison.md and migration-playbooks.md files with existing content
• Move the 4 stack templates to a reference file (keep 1 example in SKILL.md)
• Add numbered workflow steps after the decision framework
• Standardize to imperative voice throughout
• Add a "What to do if it's not working" section

Checkout your skill here: SkillzWave.ai | SpillWave We have an agentic skill installer that install skills in 14+ coding agent platforms. Check out this guide on how to improve your agentic skills.