Dispatch artifacts through multiple review skills and aggregate verdicts.
npx skills add https://github.com/chughtapan/safer-by-default --skill stamina
Please help me install this Agent Skill. Skill: stamina Source: https://github.com/chughtapan/safer-by-default/tree/main/skills/stamina Command: npx skills add https://github.com/chughtapan/safer-by-default --skill stamina
Fan-out review adapter. Routes a high-blast-radius artifact (plan or PR) through multiple heterogeneous review passes and gates on consensus. Does not review; dispatches existing review skills (/simplify, /review, /review-senior, /safer:dogfood, /codex, /security-review for PRs; /safer:dogfood + /safer:review-senior + /codex for plans) and aggregates their verdicts. N is set by the blast-radius x reversibility table in PRINCIPLES.md > Durability. Two invocation modes: --plan <sub-issue URL> and --pr <pr URL>. Use when the caller (typically /safer:orchestrate Phase 5c) has decided the artifact warrants stamina. Do NOT self-invoke from the working modality; that is Principle 5 self-polishing.
Invoke stamina with --plan or --pr and allow it to fan out reviewers according to the blast-radius rules.
High-intent search queries and answers about installing and using this skill.
You can gate consensus across multiple heterogeneous review passes by dispatching diverse review skills and aggregating their verdicts from GitHub. The adapter supports two invocation modes, --plan and --pr, to route artifacts through multiple independent reviewers.
Consensus gating aggregates per-reviewer verdicts read directly from GitHub into a single consolidated result. The dispatcher enforces independence by preventing any repeated skill and model combination across the heterogeneous review passes.
You need to fan out review skills when a caller decides an artifact warrants stamina based on the blast-radius and reversibility table. It is explicitly designed for high-blast-radius artifacts like plans or PRs that require durable consensus.
Yes, you can route a plan through multiple review passes by invoking the adapter with the --plan flag and a sub-issue URL. The dispatcher routes plans through skills like /safer:dogfood, /safer:review-senior, and /codex, aggregating their verdicts.
A key limitation is that the dispatcher does not review artifacts itself; it only aggregates existing review skills. Additionally, you should not self-invoke the adapter from the working modality, as that violates the self-polishing principle.
chughtapanRun lint, type-check, and test commands to produce a SHIP or HOLD verdict.
team-attentionConverts draft GitHub pull requests to Ready for Review state.
YaugourtRoute code diffs through the Codex CLI for review, challenge, or consult modes.
jedmamostoReview phase plans by invoking external AI CLIs and aggregating critiques into REVIEWS.md.
