templates/skills/project/fast/stack-pick/SKILL.md
Fast, evidence-based technology stack selection for one-iteration planning
npx skillsauth add ozerohax/assistagents project-fast-stack-pickInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
<when_to_use> <item importance="critical">After pulse-scan and before proto-spec finalization</item> <item importance="high">When there are 2+ realistic stack or platform options</item> </when_to_use>
<required_preload> <item>planning-approach-selection</item> <item>planning-estimation</item> <item>planning-risk-assessment</item> </required_preload>
<inputs> <required>Constraints, risks, and success criteria from previous steps</required> <optional>Current team expertise and available resources</optional> <optional>Compatibility requirements with the current landscape</optional> </inputs> <method> <step>Define 3-6 selection criteria (speed, risk, maintainability, cost, compatibility)</step> <step>Describe 1-3 realistic stack options</step> <step>Compare options against criteria and choose the primary one</step> <step>Capture trade-offs and a fallback option</step> <step>Validate that the choice supports MVP delivery within one planning iteration</step> </method><output_format>
<section>Decision criteria</section> <section>Stack options</section> <section>Chosen stack + rationale</section> <section>Trade-offs</section> <section>Fallback option</section> </output_format><quality_rules> <rule importance="critical">Stack selection is justified by criteria, not preferences</rule> <rule importance="high">Trade-offs and limitations of the chosen option are documented</rule> <rule importance="high">A fallback exists for critical risks</rule> </quality_rules>
<do_not> <item importance="critical">Do not choose a stack without explicit comparison of at least two options (when options exist)</item> <item importance="high">Do not ignore actual team expertise</item> </do_not>
testing
Visual regression (baseline diffs, UI stabilization)
testing
Design detailed reproducible test cases (Given/When/Then, one check per case); not smoke/regression checklists
development
Mobile testing (native/hybrid/web, gestures, offline, interruptions)
testing
Load testing (k6/JMeter scenarios, benchmarks)