Stats
Actions
Tags
From research-workflow
Use when a result or model rests on simplifying assumptions, approximations, fixed parameters, or regime-of-validity choices — keep an explicit running ledger of what each result depends on, so when an assumption later breaks you know exactly which conclusions die with it. Don't use for citing the source of a value (→ provenance-of-constants), recording a decision and its rationale (→ decision-log-and-commits), or quantifying the numeric error a kept assumption induces (→ uncertainty-reporting-gate).
How this skill is triggered — by the user, by Claude, or both
Slash command
/research-workflow:assumption-ledgerThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Every result carries load-bearing "this is only true if…" clauses — a linearization, a frozen parameter, an ignored term, a regime of validity, a data/instrument caveat. Left implicit, they are invisible until one quietly fails and silently invalidates a conclusion nobody re-examined. Keep them explicit: a living ledger mapping each significant result to the assumptions it stands on, so a broke...
Every result carries load-bearing "this is only true if…" clauses — a linearization, a frozen parameter, an ignored term, a regime of validity, a data/instrument caveat. Left implicit, they are invisible until one quietly fails and silently invalidates a conclusion nobody re-examined. Keep them explicit: a living ledger mapping each significant result to the assumptions it stands on, so a broken assumption triggers a known, bounded re-check instead of an undetected wrong answer.
β(r) fixed to Osipkov–Merritt, self-gravity neglected, optically thin, linear regime, Gaia DR3 completeness assumed flat) — not "standard approximations."valid for r ≪ r_t, breaks above τ ~ 1).This tracks what a result depends on. The source of a value is provenance-of-constants; the rationale for a choice is decision-log-and-commits; the numeric error a kept assumption introduces is uncertainty-reporting-gate.
provenance-of-constants — cites where a value came from; this records the conditions under which it applies.decision-log-and-commits — the decision and why; this is the standing list of what the decision assumed.uncertainty-reporting-gate — assumptions in the ledger often map directly to systematic-error terms.null-result-integrity — a broken assumption may be exactly why a result failed; record both.npx claudepluginhub drannarosen/research-workflow --plugin research-workflowGuides creation, editing, and verification of skills for AI coding agents using test-driven development with subagent scenarios. Use when authoring or debugging skills.