← ClaudeAtlas

hypothesislisted

Populate or expand a project's docs/hypotheses.md (or docs/hypotheses/ folder) — testable predictions linking theory (docs/theory.md or docs/literature.md) to evidence and the project's empirical design. Use when the user wants to draft, update, or audit hypotheses.
hsigstad/research-kit · ★ 0 · AI & Automation · score 75
Install: claude install-skill hsigstad/research-kit
# /hypothesis — Populate docs/hypotheses.md Draft or extend a project's `docs/hypotheses.md` (or `docs/hypotheses/` folder): a structured mapping of theoretical predictions to pre-existing evidence and concrete empirical tests the project can run. Each hypothesis is a single testable claim with its theory, evidence, specification, and data requirement spelled out. ## Locating the hypotheses doc Discovery order, first hit wins: 1. `$PROJ/docs/hypotheses.md` (flat file) 2. `$PROJ/docs/hypotheses/index.md` (folder mode) If neither exists, ask the user which to create. Default to flat file for <15 hypotheses, folder for 15+. ## Arguments - `/hypothesis` — infer project from the current working directory. - `/hypothesis <project-slug>` — run against a specific project under `projects/`. - `/hypothesis --audit` — do not write; report gaps, dangling refs, and orphan theories. - `/hypothesis --extend` — preserve existing entries; only append new hypotheses suggested by current evidence/literature. - `/hypothesis --update <H#> [--artifact <AN-id-or-build-path>]` — **surgical single-entry edit**. Update the status block (or evidence/test fields) of one existing hypothesis after a new analysis result. Reads only `CLAUDE.md`, `hypotheses.md`, and the optional triggering artifact. `--artifact` accepts either an AN id (preferred — e.g. `AN-019`; the skill reads `docs/analyses/an-019-*.md` to recover script + target + headline + confidence) or a raw `build/<path>` (fallback for unled