Detect contradictions, unsupported inferences, and missing cases in bounded source sets.
npx skills add https://github.com/srinitude/skills --skill logic-audit
Please help me install this Agent Skill. Skill: logic-audit Source: https://github.com/srinitude/skills/tree/main/skills/logic-audit Command: npx skills add https://github.com/srinitude/skills --skill logic-audit
This Skill helps you detect contradictions, missing links, unsupported inferences, and authority conflicts in a bounded set of sources before you make a decision or publish a conclusion.
Ask the skill to audit a bounded source set for contradictions, missing cases, and proof gaps, then report the findings with the smallest safe repair.
High-intent search queries and answers about installing and using this skill.
To detect contradictions in policies or specifications, normalize the bounded source set, trace dependencies, and test for counterexamples. This process identifies direct conflicts, scope drift, and unsupported inferences before you finalize decisions.
Logical consistency checking compares claims, evidence, and conclusions to identify missing cases and weak evidence chains. You need it for research notes when scope, time, modality, or quantifiers may drift across mixed evidence sources.
Audit a bounded source set by applying source normalization and counterexample testing to detect authority conflicts and missing cases. The process returns a bounded finding ledger that distinguishes confirmed issues from unresolved uncertainty.
The best way to repair logical contradictions is through bounded repair, which produces the smallest coherent fix while preserving intent, authority, and exact methods. It operates within explicit evidence limits to maintain source validity.
Yes, you can audit specifications to check for quantifier changes and conditional reversals. The auditing process compares requirements and claims to detect scope drift and logical conflicts across the bounded source set.
A key limitation is that logical inference analysis requires a strictly bounded source set and cannot resolve issues beyond explicit evidence limits. It distinguishes confirmed findings from unresolved uncertainty but cannot repair gaps outside the source boundaries.
magicpro97Transform raw business needs into a structured, traceable requirements specification.
aaronjmarsDiagnose failing AI skills by analyzing logs and issue trackers.
anorthamTrace call paths and search symbols to understand unfamiliar codebases.
zapabobTrain SFT+GRPO models for mathematical theorem proving with MCP/A2A integration.
