Stats
Actions
Tags
From crux
Traces causal chain of a decision in Crux graph back to root constraint via dependsOn links. Displays hierarchy from leaf to root for rationale review.
How this skill is triggered — by the user, by Claude, or both
Slash command
/crux:crux-whyThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
The user wants to understand why a decision was made. The argument "$ARGUMENTS" is a search query for a decision.
The user wants to understand why a decision was made. The argument "$ARGUMENTS" is a search query for a decision.
.claude/.crux/sessions/ in the project root (find the most recent .json file, or the one matching the current session)dependsOn linksFormat:
Decision: Use REST not GraphQL
└─ Rationale: REST is the only viable choice given team skills
└─ Constraint: Team only knows REST — GraphQL is not an option
└─ Stakeholder: Backend team lead (if present)
If multiple decisions match, show all of them. If no match is found, list all active decisions so the user can pick one.
npx claudepluginhub akashp1712/claude-marketplace --plugin cruxExposes Claude's reasoning chain as an auditable, decomposable artifact. Quick mode yields assumption inventory and weakest link; full mode adds decision branching, confidence decomposition, and falsification conditions.
Queries AI decision trails for code reviews (CRs), gates, analysis (intent/change/risk/autonomy), architecture decisions (ADR), and timelines. Audits judgments with confidence scores, processes, and context navigation.
Exposes Claude's reasoning as auditable traces with atomic claims, assumption ratings, weakest links, confidence decomposition, and falsification conditions. Triggers on 'reasoning', 'why', 'trace'.