The evidence model

The promise of this library is honest grading. The field of “thinking tools” is a small empirical core surrounded by a large practitioner ring and a weak outer ring. Most libraries flatten that into uniform confidence. This one labels which is which, on every skill and every claim.

The seven tiers

Tier	Meaning
S	Strong research - replicated experimental or meta-analytic support.
M	Moderate - real evidence, but narrower, correlational, or field-based.
P	Practitioner - widely used and defensible, without strong controlled evidence.
V	Vendor / commercial - originates from a consultancy or branded methodology.
A	Anecdotal - case reports and testimonials.
C	Conceptually plausible, under-tested - reasonable, not yet demonstrated.
X	Poor or contradictory - the evidence cuts against it (these are excluded, documented).

A “P, useful anyway, here is when not to use it” is more trustworthy than a dressed-up “S”. A practitioner-tier framework is still worth running; it is just labeled honestly.

What the grade does and does not promise

• The tier reflects the evidence for the framework’s mechanism, and the page’s dossier says explicitly what the research shows and what it does not. For example, the premortem reliably surfaces more and more-specific risks (well supported), but is not proven to improve final outcomes (a common overclaim).
• Transferred-evidence flag: most evidence comes from human-subject studies. Almost none of it tests an AI agent running the method. Where that is the case, the page says so. Treat the AI value as: the agent makes the mechanism cheap to run, enforces the structure, and produces a durable artifact.
• No laundered statistics: effect sizes and figures appear only where a primary source supports them, with the caveat the source records.

Browse by evidence

The bibliography aggregates the graded sources across the library, so a skeptic can trace any claim to its grounding. The strong-evidence core (the S-tier skills) is the anchor; the rest is honestly labeled around it.