Skip to content

pedrohcgs/claude-code-my-workflow / .claude/skills/challenge

challenge

Verified

Stress-test a finding against the choices you did not make. Enumerates the discrete forks a competent analyst could have taken (measure definition, sample filter, control set, clustering level, weighting, functional form), runs the specification grid, and reports the distribution rather than a point estimate — then attacks the identifying assumption with named, computable sensitivity statistics. Use when the user says "is this robust", "challenge this result", "specification curve", "multiverse", "how sensitive is this", "what if I'd used a different measure", "stress-test my estimate", or before a result becomes a headline claim. NOT a reviewer of prose or code — it challenges the CLAIM.

FileKindSize
SKILL.mdskill md7.6 KB
evals/marker.txtmarkdown93 B
references/fork-catalog.mdmarkdown2.0 KB
references/sensitivity-statistics.mdmarkdown2.8 KB

references/fork-catalog.md

View on GitHub
# Fork catalogue — where competent analysts diverge, by design

Use when enumerating forks for a design you have not challenged before. A fork belongs in the
grid only if you would **defend either branch** in a seminar. Padding the grid with
indefensible alternatives dilutes the fragile cells and flatters the result.

Mark each fork **`estimand`** (changes what is being estimated) or **`estimate`** (same target,
different route). Estimand forks are reported separately — averaging over them is meaningless.

---

## Every design

| Fork | Alternatives | Kind |
|---|---|---|
| Outcome scale | level · log · IHS · share · rate | `estimand` if the elasticity target differs |
| Winsorization / trimming | none · 1% · 5% | `estimate` |
| Sample window | full · balanced · excluding a crisis period | `estimand` |
| Clustering | unit · assignment level · two-way | `estimate` (inference only) |
| Weighting | unweighted · population · IPW | `estimand` |
| Missing data | complete case · imputation · indicator | `estimate` |


> **Design-specific fork sections are deliberately absent** for the causal-identification
> designs (unvetted-methods veto): populate them from your field's canonical sources after
> vetting.

## Survey / experimental

Attention and manipulation-check screens · pre-registered vs full sample · multiple-comparison
correction (none · Holm · BH) · attrition handling. Screens are **`estimand`** — they change
the population.

---

## Sizing the grid

Grid size is the product of the alternatives. Six forks with three options each is **729 fits**.

- Start with the **3–4 forks most likely to move the result** — usually measure definition,
  comparison group, and control set.
- Run those fully, then expand only if the curve is tight.
- If a full grid is infeasible, run a **fractional design** and say so — a sampled grid honestly
  labelled beats a full grid nobody ran.
- **Log what you did not run.** Silent truncation reads as "we covered everything".
finishing-a-development-branchobra/superpowersUse when implementation is complete, all tests pass, and you need to decide how to integrate the workStars: 322kUpdated Claude CodeCodexVerifiedInstall
systematic-debuggingobra/superpowersUse when encountering any bug, test failure, or unexpected behavior, before proposing fixesStars: 322kUpdated Claude CodeCodexVerifiedInstall
test-driven-developmentobra/superpowersUse when implementing any feature or bugfix, before writing implementation codeStars: 322kUpdated Claude CodeCodexVerifiedInstall
skill-creatoranthropics/skillsCreate new skills, modify and improve existing skills, and measure skill performance. Use when users want to create a skill from scratch, edit, or optimize an existing skill, run evals to test a skill, benchmark skill performance with variance analysis, or optimize a skill's description for better triggering accuracy.Stars: 198kUpdated Claude CodeCodexVerifiedInstall