Designs and executes compatibility-preserving migrations across APIs, schemas, frameworks, dependencies, protocols, storage, and infrastructure with baselines, coexistence, staged cutover, rollback or forward-fix paths, and explicit removal criteria. Use for deprecations, upgrades, data moves, contract changes, and replacing live implementations. Not for a local refactor with no external state or compatibility boundary.
インストールする前に、エージェントに与えられる指示の中身を確認できます。
Treat compatibility, state, and reversibility as first-class requirements.
Identify the forcing function, affected producers and consumers, stored state, support window, blast radius, and cost of leaving the old path in place. Reject a migration whose benefit does not justify its operational risk.
Before changing implementation, capture current contracts and representative behavior: tests, state snapshots, traffic shape, performance, error semantics, and rollback prerequisites. Existing consumers are part of the contract even when their behavior is undocumented.
Choose a pattern from migration-patterns.md: expand/contract, adapter, dual-read/write, shadow, strangler, or versioned contract. Define phases, ownership, observability, data reconciliation, entry and exit criteria, and the exact point of no return.
Prefer additive steps:
For each phase, define rollback when reversible and a forward-fix when rollback would corrupt or discard newer state. Never label an untested rollback as a rollback plan.
Prove old-state/new-code and, when relevant, new-state/old-code compatibility.
Check idempotency, interruption recovery, reconciliation, mixed-version
operation, and empty-plan/no-diff behavior. Pair production cutover with
release-engineering and data modeling with database-design.
The migration has phase-specific evidence, measurable cutover and abort gates, an owned recovery path, and explicit cleanup criteria; no destructive removal occurs while supported consumers or unreconciled state remain.
まだレビューはありません。使ってみた感想をお寄せください。
概要と使いどころ
Designs and runs reproducible evaluations for AI agents, prompts, tools, skills, and model-backed workflows using realistic datasets, isolated baselines, objective assertions, rubric grading, trajectory analysis, cost/latency tracking, and regression comparison. Use when measuring agent quality, optimizing skill triggering, comparing prompts or models, or gating an AI feature release. Not for ordinary deterministic unit tests.
日本語の概要は準備中です。原文の説明を表示しています。
Designs or reviews HTTP, REST, GraphQL, RPC, CLI, webhook, event, and service interfaces with explicit inputs, outputs, errors, compatibility, idempotency, pagination, authentication, versioning, and observability. Use when introducing or changing an API or cross-component contract. Not for internal implementation details with no boundary or for debugging one API failure; use root-cause-debugging there.
日本語の概要は準備中です。原文の説明を表示しています。
Reviews an existing codebase for structural friction, unclear ownership, leaky or shallow interfaces, excessive coupling, misplaced state, poor testability, and risky dependency direction, then prioritizes evidence-backed improvement candidates. Use for architecture audits, modularization, modernization, or recurring cross-cutting change pain. Not for designing one new interface, simplifying a local function, or fixing a reproduced bug.
日本語の概要は準備中です。原文の説明を表示しています。
Validates a running application, CLI, API, service, or generated artifact as a user or operator against a prewritten observable behavior contract while remaining source-blind. Use for acceptance checks, runtime proof, anti-fake probes, release smoke tests, or an independent companion to code review. Not for source-quality findings, root-cause diagnosis, or visual design judgment outside the contract.
日本語の概要は準備中です。原文の説明を表示しています。
Closes a completed development branch by checking the final diff and proof, presenting merge, pull-request, keep, or discard options, and cleaning up only after the user or repository workflow chooses a path. Use when feature work is complete and the branch must be integrated or retired. Not for claiming a feature is complete before verification.
日本語の概要は準備中です。原文の説明を表示しています。
Tests a real web user flow with a browser by asserting semantic behavior, network and loading states, keyboard access, responsive layouts, and stable visual evidence. Use for browser bugs, end-to-end UI behavior, responsive or accessibility checks, and screenshot baselines. Not for static source review without a browser or for backend-only tests.
日本語の概要は準備中です。原文の説明を表示しています。