Random Frameworks

Quick facts

Mechanism: Generates three off-pattern framework analyses of a topic by drawing frameworks at random, applying each regardless of fit, and harvesting the non-obvious angles the unexpected lenses expose.

Produces: applied-framework-set

Evidence: C · Domain: Meta Thinking and Reflection

Use when a framing feels frozen or fixated and the goal is to break out of it with deliberately unranked methods rather than the obvious best-fit ones. Use when relevance-ranked selection would only reinforce the stuck view and a random draw is wanted to surface what a fitted analysis would miss.

When a framing is stuck, the frameworks you would naturally reach for tend to reinforce the stuck view. This skill does the opposite: it draws three frameworks at random from the library, ignoring fit on purpose, and applies each to your topic to force lenses the situation would never summon. The value is exactly what relevance ranking suppresses. The deliberate randomness, the same logic as a random-stimulus creativity move, dislodges a frozen framing and surfaces angles a fitted analysis would miss.

When to Use

• A framing feels frozen or fixated, and the obvious frameworks would only confirm it.
• You want off-pattern lenses to surface non-obvious angles, not the best-fit analysis.
• You are exploring, early, and want breadth and surprise over precision.

When NOT to Use

• You actually want the right frameworks for a real, stakes-bearing decision. Use think-framework-advisor (it routes to the fewest fitting moves) or think-top3 (it ranks and applies the most relevant three). Applying three ill-fitting frameworks to a high-stakes, irreversible call is worse than noise.
• You are not stuck or fixated. With no frozen framing to dislodge, a random draw is overhead; route to a fit-first path.
• A single random-stimulus move would do. Use think-far-analogy-ideation (the library’s home for random stimulus); this skill is the meta-rotation of three whole frameworks, not one stimulus. For flipping a single default premise, use think-assumption-reversal.
• You want a structured rotation of fixed lenses within one method (facts, upside, risk, intuition, alternatives, process). That is think-parallel-perspectives-review, which rotates a curated lens set inside one method; this skill draws whole different frameworks at random.
• Not a thinking task (lookup, drafting, coding): redirect.

Instructions

When asked to apply random frameworks to a topic, follow these steps:

1. Parse the topic. Restate the situation in one or two sentences. If the input is under about 15 words or carries no concrete signal, ask one clarifying question, then proceed.
2. Run the shared engine in RANDOM mode. Follow references/engine.md: read the corpus, draw three frameworks uniformly at random without replacement (seeded only if the user supplies a seed), and apply each to the topic so it emits that framework’s real artifact, flagging poor structural fit rather than swapping it out.
3. Harvest the surprises. Name the one to three non-obvious angles the random draw exposed that an obvious analysis would have missed. State plainly that the set is a fixation-breaker, not a fitted recommendation.

Output Format

Use the template in references/TEMPLATE.md. The deliverable is the three randomly drawn frameworks (with the seed or fresh-draw noted), the three filled artifacts, and the harvest of non-obvious angles, not a prose essay and not a recommendation to act.

Quality Checklist

Before finalizing, verify:

• Exactly three frameworks, drawn at random (seed or fresh draw stated), each name existing in the recommendable corpus.
• Selection ignored relevance on purpose; frameworks were not quietly fitted to the topic.
• Each framework was applied, emitting its real artifact; poor structural fit was flagged, not swapped out.
• The harvest names the non-obvious angles the draw exposed and does not present the set as a fitted recommendation.
• No tier inflation; the random set carries no decision authority.

Evidence

Tier C (conceptually plausible, under-tested). Random-stimulus and forced-connection moves are a recognized creativity practice for breaking fixation, but the meta-rotation of three whole frameworks at random is not measured, in humans or in AI use. Treat the output as a fixation-breaker, never as a fitted analysis. Full grading and caveats: evidence/dossier.md.

Examples

See references/EXAMPLE.md for a completed run that draws, applies, and harvests three random frameworks on a real topic.

Deep dive: worked example

A full worked run (the shared Northwind scenario)

Example run: think-random-frameworks

A compressed worked run. The point is that the draw ignores fit; the artifacts are abbreviated to show the shape and the harvest.

Topic

“We keep circling the same framing on onboarding retention: every idea is another tweak to the welcome email sequence, and we cannot get unstuck.”

Draw

unseeded (fresh draw) - selection ignored relevance on purpose.

#	Framework	Tier	Structural fit
1	think-iceberg-model	P	good (systemic, but not the obvious reach)
2	think-fermi-estimation	M/P	poor - flagged, applied anyway
3	think-backcasting	P	partial

Applied artifacts (abbreviated)

1. think-iceberg-model

Events: users drop off after day 2. Patterns: drop-off is concentrated in users who never completed a first core action. Structures: onboarding optimizes for email opens, not first-value. Mental models: the team believes onboarding = the email sequence. The lens relocates the problem from “the emails” to “time-to-first-value”, which the welcome-email framing hides.

2. think-fermi-estimation (poor fit, applied as a fixation-breaker)

Rough order-of-magnitude: of N new signups, ~X reach the core action, ~Y of those retain. Even abbreviated, the estimate exposes that email open-rate tweaks move a number two stages upstream of retention. Exploratory only, not a measured model.

3. think-backcasting

Work backward from “users reliably retain”: the precondition is first-value within session one, which requires removing a setup step, not improving an email. The chain lands on a product change, not a messaging change.

Harvest (non-obvious angles)

• All three off-fit lenses independently relocated the problem away from the welcome email and toward time-to-first-value, the exact thing the frozen framing suppressed.
• The Fermi lens (a poor structural fit) was the one that made the “emails are two stages upstream” point most vividly.

This is a fixation-breaker, not a fitted analysis. To actually decide the retention fix, route to think-framework-advisor or think-top3 with the reframed problem.

Grounding: the full evidence dossier

What the research does and does not show, with graded sources

Evidence dossier: think-random-frameworks

What it is

A meta-skill that draws three frameworks at random from the shipped corpus, applies each to the topic regardless of fit, and harvests the non-obvious angles the off-pattern lenses expose. It is an applicator over the corpus, not a thinking method, and it is the deliberate inverse of relevance ranking.

The mechanism (the durable move)

Forced, unfitted variation. The skill removes the user’s (and the model’s) natural fit-selection, which under fixation reliably reaches for framings that confirm the stuck view. Drawing whole frameworks at random and applying them forces the situation through lenses it would never summon. The move is the same logic as a random-stimulus creativity technique, lifted from single stimulus to whole-framework rotation.

What the evidence says (honest grade: C)

Random-stimulus and forced-connection techniques are a long-standing, practitioner-grade creativity practice for breaking fixation; the underlying idea (deliberate, irrelevant input dislodges a frozen frame) is well attested as practice, weakly measured as effect. But the specific contract here - drawing three whole frameworks at random and applying each in full - is novel and unmeasured, in humans or in AI use. There is no study on N=3, on whole-framework vs single-stimulus rotation, or on whether the harvested angles outperform a directed reframe.

Honest tier: C (conceptually plausible, under-tested). The output is explicitly a fixation-breaker, never a fitted analysis; its artifacts carry no decision authority and the underlying frameworks’ tiers are carried, never inflated.

When it misleads / when NOT to use

• On a real, stakes-bearing, irreversible decision: applying ill-fitting frameworks is worse than noise; use a fit-first path.
• When the user is not actually stuck: a random draw is pure overhead.
• Reading the random artifacts as recommendations: they are prompts to dislodge a framing, not conclusions to act on. Presenting them as analysis is the failure mode.

Lineage and who to read

The lineage is the random-stimulus and forced-connection family of lateral-thinking and creativity practice (deliberately introducing unrelated input to break a fixed pattern). The library already ships single-stimulus moves (far-analogy ideation, assumption reversal); this skill is the meta-level rotation of whole frameworks, walled against those single-move siblings and against the fixed-lens rotation of parallel-perspectives-review.

Relation to the catalog fold

The framework catalog folds the single-stimulus forced-connection method into far-analogy-ideation (registry slug forced-connections, status fold). This skill deliberately does NOT re-introduce that folded method, and the distinction is the reason it can ship without overturning the fold. The folded method injects ONE unrelated stimulus and is a mode of far-analogy ideation. This skill rotates THREE whole frameworks at random and applies each, where each framework is a complete analytical apparatus that emits its own artifact. Importing a full structured procedure is distinct in kind from a single provocation, so this ships as an applicator meta-skill (no framework registry entry of its own), not as a re-add of the folded single-stimulus method. The registry entry for forced-connections records the same distinction.