Generates N independent reasoning paths for high-stakes multi-step inference and takes majority vote to surface consistent answers, avoiding single-chain errors.
npx claudepluginhub repozy/superpowers-optimizedThis skill uses the workspace's default tool permissions.
A structured reasoning technique based on the Self-Consistency method (Wang et al., ICLR 2023).
Creates isolated Git worktrees for feature branches with prioritized directory selection, gitignore safety checks, auto project setup for Node/Python/Rust/Go, and baseline verification.
Executes implementation plans in current session by dispatching fresh subagents per independent task, with two-stage reviews: spec compliance then code quality.
Dispatches parallel agents to independently tackle 2+ tasks like separate test failures or subsystems without shared state or dependencies.
A structured reasoning technique based on the Self-Consistency method (Wang et al., ICLR 2023).
Core idea: Complex problems often have multiple valid paths to the correct answer. Incorrect reasoning, even when confident-sounding, tends to scatter across different wrong answers. By generating N independent reasoning paths and taking majority vote, we reliably surface the correct answer and get a built-in confidence signal for free.
This skill is invoked internally by:
It fires when:
Scale paths to difficulty:
| Problem Type | Paths |
|---|---|
| Binary verification (does this evidence prove the claim?) | 3 paths |
| Root cause diagnosis with 2-3 candidates | 5 paths |
| Complex multi-factor diagnosis or high-stakes verification | 7 paths |
Default: 5 paths. Research shows gains plateau quickly — 5 captures most of the benefit of 40.
Produce each path independently — don't let earlier paths contaminate later ones. Vary your approach deliberately:
Each path must end with a clearly parsed final answer.
Diversity is the whole point. Paths that all use the same approach just give you one answer repeated — that's not self-consistency, it's greedy decoding in disguise.
Collect the final answers from all N paths. The most frequent answer wins.
Compute confidence:
Do not show all paths to the user. The process is internal. Surface only the aggregated result:
**[Diagnosis/Verdict]**: [the majority-vote answer]
**Confidence**: [X/N paths agree] [high/moderate/low]
[Only if confidence < 80%]: Brief note on minority conclusion and the key divergence point.
systematic-debugging — invokes SC during root cause hypothesis testingverification-before-completion — invokes SC during evidence evaluation