Designing The Analysis
Turns an approved research question into a pre-registration-ready, step-by-step analysis plan
Test report
- Verdict
- Tested · Works
- Score
- Tested
- Jul 31, 2026
- Environment
- Claude Code 2.x (agent harness)
- Upstream re-checked
- Aug 10, 2026 · 92b535e
I fetched SKILL.md and confirmed the frontmatter parses with name+description and no scripts or security smells; the three sibling skills its body hands off to (preregistering-analysis, subagent-driven-analysis, executing-analysis) each returned HTTP 200. For output I designed one task — an analysis plan for "does the new onboarding flow raise 7-day retention?" — and wrote a baseline plan without the skill then a skill-following plan. Concrete differences: the baseline's decision rule was result-dependent ("if the new flow shows higher retention, recommend rolling it out") while the skill forced a rule fixed in advance ("variant OR>1 with 95% CI lower bound >1.0"); the baseline said "adjust for confounders as needed" whereas the skill named channel + signup-day and how each is handled; the baseline named no test and had zero power analysis, while the skill specified the exact logistic formula plus a +3pp MDE, N≈7,400/arm, alpha and power, and added simulated-data validation, immutable-raw dataflow, checkbox steps and a prereg handoff header the baseline lacked entirely.
Scored on four weighted criteria — install, triggering, output vs. baseline, docs. How scoring works
- Installs cleanly 5/5
- Triggers reliably 5/5
- Output vs. baseline 8/10
- Docs & honesty 5/5
What Designing The Analysis does
A methodology skill that converts an approved research question into a concrete analysis plan: exact datasets and variables, the named model or test, a power/sample-size justification, a confound-handling plan, and a decision rule fixed before any outcome is seen. Triggers when you have a settled research question and need to design the analysis before touching outcome data or fitting a model. Output is broken into bite-sized validated steps and hands off to the preregistering-analysis skill.
How to install Designing The Analysis
git clone --depth 1 https://github.com/K-Dense-AI/science-superpowers.git /tmp/designing-the-analysis-src
mkdir -p ~/.claude/skills
cp -R /tmp/designing-the-analysis-src/skills/designing-the-analysis ~/.claude/skills/designing-the-analysis
# Pure prompt/methodology skill: no scripts, no third-party deps (POSIX shell only).
# The plan body references sibling skills in the same repo (preregistering-analysis,
# subagent-driven-analysis, executing-analysis) for the full workflow; install those too
# by copying the other dirs under /tmp/designing-the-analysis-src/skills/ if you want the chain.
# Plans are written to docs/science-superpowers/plans/YYYY-MM-DD-<topic>.md in your project.
Skills live in ~/.claude/skills/ (global) or .claude/skills/
(per-project). Restart Claude Code after installing.
Commands — how to trigger Designing The Analysis
-
/designing-the-analysisTurns an approved research question into a pre-registration-ready, step-by-step analysis plan
It also activates on plain-language prompts like these:
-
Design an analysis plan for this research question -
Pre-register the analysis before I see the data -
Plan the statistical approach for this study
Frequently asked questions
- Is the Designing The Analysis skill free?
- Yes. The skill itself is free from K-Dense-AI/science-superpowers. SkillProof publishes the install command and an independent test verdict at no cost.
- Does Designing The Analysis work with Claude Code?
- We tested it with Claude Code 2.x (agent harness) on Jul 31, 2026. Verdict: Tested · Works. I fetched SKILL.md and confirmed the frontmatter parses with name+description and no scripts or security smells; the three sibling skills its body hands off to (preregistering-analysis, subagent-driven-analysis, executing-analysis) each returned HTTP 200. For output I designed one task — an analysis plan for "does the new onboarding flow raise 7-day retention?" — and wrote a baseline plan without the skill then a skill-following plan. Concrete differences: the baseline's decision rule was result-dependent ("if the new flow shows higher retention, recommend rolling it out") while the skill forced a rule fixed in advance ("variant OR>1 with 95% CI lower bound >1.0"); the baseline said "adjust for confounders as needed" whereas the skill named channel + signup-day and how each is handled; the baseline named no test and had zero power analysis, while the skill specified the exact logistic formula plus a +3pp MDE, N≈7,400/arm, alpha and power, and added simulated-data validation, immutable-raw dataflow, checkbox steps and a prereg handoff header the baseline lacked entirely.
- What is the Designing The Analysis SkillProof Score?
- 9.2/10 — installs cleanly 5/5, triggers reliably 5/5, output vs. baseline 8/10, docs & honesty 5/5.
- How do I install Designing The Analysis?
- Copy the install command from this page, run it in your terminal, and restart Claude Code. Skills live in ~/.claude/skills/ (global) or .claude/skills/ inside a project.
- Can I use Designing The Analysis with Cursor, Copilot, Gemini CLI, Codex or other AI tools?
- The SKILL.md format is native to Claude (Claude Code, Desktop, claude.ai). The instructions inside adapt to other assistants: Cursor rules, GitHub Copilot instructions, Windsurf rules, Custom GPTs, AGENTS.md for OpenAI Codex, and GEMINI.md for Google Gemini CLI — our conversion guides cover each, and the free converter on the tools page does the wrapping for you.