Enumerate assumptions embedded in a validator/test/sandbox, including data-generation, measurement, simulator, oracle, metric, initialization, and acceptance assumptions.
Installs into .claude/skills of the current project.
Are you the author of Enumerate Validator Assumptions?
Add the live security badge to your README. It updates with every re-scan.
[](https://www.skillsdirectory.com/skills/yogsoth-ai-enumerate-validator-assumptions)
---
name: enumerate-validator-assumptions
description: "Enumerate assumptions embedded in a validator/test/sandbox, including data-generation, measurement, simulator, oracle, metric, initialization, and acceptance assumptions."
---
# enumerate-validator-assumptions
## Purpose
Enumerate assumptions embedded in a validator, benchmark, sandbox, or test.
## Input contract
```yaml
required: [validator_artifact, target_claim]
optional: [data_generation, measurement, simulator, oracle, metric, initialization, acceptance_rule]
constraints: [assumptions must be classified by source and role]
```
## Procedure
1. Inspect data generation, measurement, simulator, oracle, metric, initialization, and acceptance.
2. Record each assumption and its relation to the target.
3. Mark evidence, uncertainty, and potential circularity.
If validator assumptions and dependencies are explicit, consider `build-noncircularity-matrix` as the next tactic.
## Output contract
```yaml
produces: [validator_assumption_inventory, dependency_links, circularity_candidates]
delta_fields: [findings, evidence_updates, uncertainties, open_questions]
```
## Quality gates
- No assumption is omitted merely because it is conventional.
## Failure and counterexamples
An unstated assumption remains unknown, not independent.
## Provenance map
- resolved: circular-validation-audit
