Claude Skills by chiruu12

Get a candidate judged by a panel of independent, read-only reviewer agents - each pinned to the model that is actually good at its lens - and act on their reports. Use before a high-stakes merge, on a plan you are about to commit to, on a commit range you did not write, or to audit a repo. The engine runs the panel and hands you one report per reviewer plus a unified one; it computes no verdict. This playbook is the judgment: which panel to reach for, how to write a lens, and how to read sev...

Compare AI coding strategies on a real task with Marshal: run the same goal through several configured clients (e.g. a premium model vs a cheaper one) and get a measured cost/latency/outcome comparison. Use when you want evidence for which provider/model to route a kind of work to, or to prove what a cheaper strategy actually saved. Honest by construction - it measures, it never guesses.

Run a full development cycle on the fleet: decompose a goal into independent worker jobs, spawn them in isolated worktrees, verify what came back, gate the risky diffs through a review panel, integrate one at a time, and record every outcome including the rejections. Use when a coding goal has parallelisable parts you would otherwise do serially. Trigger on "delegate this", "fan this out", "run the fleet", "build this with marshal", "parallelise this work".

Drive a fleet of headless agents through Marshal's MCP server: decompose a goal into independent tasks (implementation, research, review, audit, summarise), run them in parallel in isolated git worktrees, collect each run's product (diff or text), and integrate the good diffs. Use when you have multi-part work to delegate to worker agents (Cursor, OpenCode, Codex, Antigravity, Claude Code) instead of doing it all yourself. The engine is mechanism; this playbook is the judgment - decomposition...

Run an adversarial review panel on a plan, a diff, or a commit range, and decide what to do with what comes back. A thin, opinionated front end to marshal-adversarial-review: how to invoke it, how to route lenses across models, which fields mean a lens is missing rather than satisfied, and why a unanimous panel is still not a decision. Use before a risky integrate, before opening a PR on a repo you do not own, or on a plan you are about to commit to. Trigger on "run a panel", "adversarial rev...

Reach a consensus on the APPROACH before committing a fleet to build it. Spawn several biased, independent solvers that each propose a plan for the same question, then an independent judge that converges them into one concrete plan (or narrows the question and re-runs). Use when the approach is contested - multiple valid designs, ambiguous scope, or a costly wrong turn - not for obvious work. The engine runs the agents; this playbook is the judgment. Hand the converged plan to marshal-orchest...

Gate a candidate diff behind an independent, multi-reviewer consensus before integrating it. Spawn several biased reviewers (architect, quality, tests) that each judge the SAME diff in isolation, then apply a fixed truth table - integrate only when no reviewer rejects and at least one approves. Use when a single-pass review (marshal-orchestrate step 4) is not enough: a high-stakes merge, or when you want a quorum instead of one opinion. The engine runs the agents; this playbook is the judgment.

Author and run declarative YAML workflows on the Marshal fleet - reusable orchestration recipes (fan a goal out across clients, collect the diffs, then merge the good ones) that you run by name instead of re-planning each time. Use when the same multi-agent shape recurs (review, compare, fix-across-clients). For a one-off decomposition, use marshal-orchestrate instead. The engine runs the recipe by sequencing safe primitives; the judgment - which recipe, and which diff to keep - stays with you.