yogsoth-ai Hypothesis Formulation
Transform insights and gaps into structured testable hypotheses — answering "how do we turn an insight into a testable hypothesis?"
HARD-GATE
Not met → stop, recommend completing gap-prioritization or upstream work first.
Campaign Goal
Transform a vague "there's a gap here" or "this phenomenon is interesting" into a precise, falsifiable hypothesis with boundary conditions. The output is not a "solution" but a "testable proposition."
Strategy Selection
| Strategy | When to Use | Reasoning direction |
|---|---|---|
| deductive-hypothesis-generation | The domain has mature theory, and the gap is a deviation between theoretical prediction and reality | Theory → prediction |
| inductive-hypothesis-generation | The domain lacks theory but has abundant empirical observations or data patterns | Data → regularity |
| abductive-hypothesis-generation | An anomalous phenomenon that existing theory cannot explain has been observed | Anomaly → best explanation |
| competing-hypothesis-construction | Need to maintain openness and avoid confirmation bias | Multiple explanations in parallel |
| hypothesis-operationalization | A directional hypothesis exists and needs to be formalized into a testable form | Vague → precise |
CC selects autonomously based on input characteristics (whether there is theoretical support, whether there is anomalous data, hypothesis maturity). Can be combined serially.
Budget Gate
| Tier | Hypothesis yield | Theory/mechanism | Falsifiability | Competing hypotheses |
|---|---|---|---|---|
| S | ≥2 structured hypotheses | ≥2 relevant theories | 1 falsification scenario per hypothesis | Optional |
| M | ≥3 structured hypotheses | ≥3 theories + ≥5 mechanisms | ≥1 scenario + boundary conditions per hypothesis | ≥2 competing hypotheses |
| L | ≥5 structured hypotheses | ≥5 theories + ≥8 mechanisms | Complete falsifiability audit | ≥3 competing hypotheses + discriminating predictions |
Hypothesis Structure (standard output format)
Each hypothesis must contain:
- Statement: If X then Y (condition-result form)
- Variables: independent, dependent, control, and moderator variables
- Mechanism: why X causes Y (causal chain)
- Boundary conditions: the preconditions under which the hypothesis holds
- Falsification: what observation would refute this hypothesis
- Measurability: how to measure it (operational definitions)
Context Management
- Invoke context-init at campaign start
- Invoke context-checkpoint after each strategy completes (hard constraint)
- All outputs accumulate in a single campaign-scoped context file
Minimum Yield
Each campaign run must produce:
- ≥2 complete structured hypotheses (including all 6 components)
- A falsifiability argument for each hypothesis
- Relationship notes between hypotheses (complementary/competing/independent)
- A recommended next direction (which hypothesis is most worth further developing into a research question)
Available Strategies
Optional, no fixed order; the final leaf is always a sop.
| Strategy | When to use |
|---|---|
| abductive-hypothesis-generation | Strategy: Inference to the best explanation in the face of anomalies |
| competing-hypothesis-construction | Strategy: Construct multiple competing hypotheses for the same phenomenon |
| deductive-hypothesis-generation | Strategy: deduce hypotheses from existing theory |
| hypothesis-operationalization | Strategy: refine a working hypothesis into a precise, testable form |
| inductive-hypothesis-generation | Strategy: induce and distill hypotheses from data/observations |
Available SOPs
Optional, no fixed order; the final leaf is always a sop.
| SOP | When to use |
|---|---|
| context-checkpoint | Append research process and results to the current Phase's context file. Covers both process and results with genuine substance. Use this skill at plan-designated checkpoint points — typically after each strategy completes or at key decision nodes within a research Phase. |
| context-init | Create a new context file for a research Phase. Called once at Phase start to initialize the file that subsequent context-checkpoint calls will append to. Use this skill whenever a new research Phase begins and a fresh context file is needed. |
| hypothesis-formation-quality-gate-check | Shared SOP: General quality-gate check (format completeness, logical consistency) |
| hypothesis-formation-saturation-detection | Shared SOP: judge whether the current activity has reached information saturation |
| hypothesis-synthesis | SOP: synthesize all intermediate products into a final structured hypothesis set |
Available Campaigns
Optional, no fixed order; the final leaf is always a sop.
| Campaign | When to use |
|---|---|
| research-question | Campaign: Refine hypotheses into precise, framed research questions |