challengelisted

Two-phase challenge that judges whether a discovered problem is real and whether a proposed idea is usable. Phase 1 loops subagent verification and advisor judgment over evidence (OUTCOME.md + parallel subagents) to self-resolve design-tree branches. It asks the user only the irreversible branches that remain and proceeds on stated assumptions for the rest. Phase 2 spawns two critic-design subagents (internal attack / OUTCOME.md attack) as devil's advocate input. The verdict leads the output as a simple GO / NO-GO. Do NOT use for code review findings (use the audit workflow) or outcome assertion (use /assert which has built-in adversarial testing).
thkt/dotclaude · ★ 12 · AI & Automation · score 78

Install: claude install-skill thkt/dotclaude

# /challenge - GO / NO-GO verdict on a proposal Judge the proposal in two phases, so the next decision starts from a verified GO / NO-GO. ## Input `$ARGUMENTS` carries the target. It may be a proposal file path or a description. If empty, stop and ask the user to specify the target; do not infer it from the conversation. When multi-line, the first line is the target title. ## Phase 1: Grill Verify the proposal from evidence, then return only the unsettled questions to the user. ### Step 1: Settle the questions with evidence Every judgment below is decided by the table, never by advisor confidence. | Subject | Condition | Treatment when it holds | Treatment when it does not | | --------------------- | -------------------------------------------------------------------------------------------------------- | ------------------------------------------- | -------------------------------------------------------------------------------- | | A question | Evidence settles it to one answer. One needing a choice, such as priority or scope, does not settle | subagents verify its answer in parallel | Send it unverified to the unsettled pile | | A verified fact | The targeted state already holds, or it contradicts the proposal. advisor o