1,280 件 ・ 関連度順
概要と使いどころ
設計文書、分析レポート、報告書などを批判的な視点で徹底レビューするスキル。 6つの異なる立場（開発者、PM、顧客、QA、セキュリティ、運用）のペルソナを持つサブエージェントが並列でレビューを実行し、 「本当にそうか？」「根拠は何か？」「テストできるか？」「運用できるか？」という視点で 曖昧さ、根拠不足、論理飛躍、テスト不能、セキュリティリスク、運用懸念を検出する。 Use when reviewing design documents, analysis reports, incident reports, or any document requiring rigorous validation of claims and evidence.
Judge performance or resource claims on real hardware: benchmark design or review, latency, throughput, memory or headroom numbers, and accept-or-reject decisions on optimizations. Not for refactors with no resource claim.
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。
Work the verification ladder of a research run with rs_verify — read the ladder report before claiming anything is checked, run the two file rungs (r1-record, r2-locate) that need no reader, record the r3-support verdicts and strength verdicts that only a reader can make, tell a miss from an inconclusive, and know what a clean build will still refuse. Use when a document is written and about to be called checked, when a claim's quote does not match its source, or when a gate or build answers with an integrity refusal.
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。
Read the Free AI Credits directory — enumerate currently-claimable free AI credit offers, fetch per-offer terms and claim steps, and interpret the site's trust labels — using only static HTTPS GETs.
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。
Gauntlet for absolute claims (always/never/guaranteed/optimal); pressure-test, then refine with explicit boundaries. Use when users ask to prove or disprove strong certainty claims, request devil's-advocate challenge rounds, or want the /tw-prove-it gauntlet to run all 10 rounds continuously by default.
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。
Turns sustainability or ESG reports the user provides into a DRAFT question pack for reviewers: every quantitative disclosure as stated with unit, period, boundary and page, year-over-year changes computed only from the stated figures with the arithmetic shown, targets with stated and computed progress, restatements and inconsistencies, gaps and specific questions to the report owner. No assurance, accuracy or conformance verdict. Use when the user asks to "prepare review questions on our sustainability report", "what changed year on year in this ESG report", "check the figures in this report before it goes to the board", "find the gaps in this ESG disclosure" or "build a question pack for the report owner". Do not use for reviewing marketing text against environmental-claim rules, use green-claims-review instead. Drafts for human review; never approves, authorises or signs off.
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。
Turns a control list and an audit scope into draft evidence requests grouped by control owner, each with the control as stated, the audit period, evidence examples (commonly requested) and a proposed due date derived from the fieldwork dates. Never judges whether evidence is sufficient, never selects the sample and never sends a request; the auditor decides. Use when the user asks to "prepare the evidence requests", "draft the PBC list", "build the evidence request pack for the audit" or "what do we need to ask each control owner for". Do not use for document requests and interview questions built from an audit plan, use audit-prep-pack instead; to check claims in a report against sources, use claims-evidence-map. Drafts for human review; never approves, authorises or signs off.
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。
Drafts one anonymised DRAFT case study (situation, approach, outcome as evidenced, optional lessons) from engagement notes, status reports, closing reports, emails and metrics extracts the user supplies, with a fact trace to the notes, an evidence grade on every result, quotes only where the user supplied them with speaker role and permission, an anonymisation log and a confidentiality checklist that must be completed before publication. Use when the user asks to "write a case study from these notes", "turn this engagement into a customer story", "draft a success story for the website", "anonymise this project write-up" or "prepare a reference case for the proposal". Do not use for redacting identifiers from an existing document, use document-deidentification-pass instead; for checking whether a document's claims are substantiated, use claims-evidence-map. Drafts for human review; never approves, authorises or signs off.
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。
Turns a set of questions (form exports, helpdesk logs, meeting Q and A, chat threads, a plain list) plus the source documents the user supplies into a DRAFT FAQ grouped by theme: every answer quoted or paraphrased from a named source with a reference, verbatim quotes for dates, figures, eligibility and obligations, UNKNOWN where the sources are silent, a source register with precedence, a conflicts table and a routing list of unanswered questions to owners. Use when the user asks to "build an FAQ from these documents", "answer these questions from the policy", "turn the Q and A into an FAQ page", "group these questions and answer them from the guide" or "what do our documents say about these questions". Do not use for the manager FAQ of a changed HR policy, use policy-change-briefing instead; for checking whether a document's own claims are evidenced, use claims-evidence-map. Drafts for human review; never approves, authorises or signs off.
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。
Drafts an architecture or design decision record (ADR) from the discussion notes, design review minutes, chat threads and diagram descriptions the user provides: context, decision drivers, options considered with arguments as stated, the decision as the group worded it, consequences and a review date, with every factual claim tagged evidenced, asserted or UNKNOWN so unsupported claims are marked for the author to substantiate. Use when the user asks to "write an ADR", "document the architecture decision", "record why we chose this design", "turn these design review notes into a decision record" or "draft the design decision log entry". Do not use for a business or governance decision with no technical design, use decision-memo-builder instead; for the change ticket that implements the decision, use change-request-pack. Drafts for human review; never approves, authorises or signs off.
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。
- Find leaked or mentioned selectors circulating in pastes, leak forums, Telegram channels and dump markets, and judge whether a claimed leak is genuine or a recycled combolist. Use when checking whether a name, email, domain or credential is circulating, verifying a breach claim made against your organisation, or setting up ongoing leak monitoring
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。
Use when the user asks to "turn this idea into posts for every platform", "write the X thread / LinkedIn post / 小红书 note", or "spec the carousel slides"; turns one idea into N platform-native ready-to-paste packages — post/thread, caption, Threads text, 小红书 note, link post with first-comment placement — each adapted per the dated platform norm card (never verbatim cross-posting), every product claim held to approved claims-ledger wording or flagged [needs source], every variant tagged {formula | hook family | CTA type | signal optimized}, plus a carousel slide-spec mode (intro/content/outro slide roles, per-element char budgets, 小红书 3:4 / IG 4:5 / LinkedIn document PDF artboards, text-safe-zone + alt-text checklist). Not for creator deliverable briefs — use brief-generator; not for repurposing existing assets with paid boost — use content-amplifier. 社媒文案/一稿多发改写/小红书笔记/轮播图脚本
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。
Use when the user asks to "build a story bank", "collect our origin and customer stories", or "assemble reusable proof stories for the message"; assembles reusable narrative units — origin, founder, customer, transformation, and proof stories — each tagged to a claims-ledger ID and a message-house pillar, with every proof labeled Measured / User-provided / [needs source]. Not for authoring the message house or pillars — use message-system-architect; not for brand voice or naming rules — use brand-language-codifier; not for finished long-form prose — use content-writer; not for adjudicating whether a proof is true — use offer-claims-registry. 品牌故事库/起源客户转化/证据故事单元
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。
Use when the user asks to "build a message house", "write a PR-FAQ for our launch", or "define the launch narrative and value pillars"; derives from the positioning canvas a message house — tagline, one-liner, three value pillars, per-persona proof points (each labeled Measured / User-provided / [needs source]) — plus a working-backwards PR-FAQ narrative spine (launch-day tense, empty-chair test, five external + five internal FAQs) and per-channel message angle packs (angles, not finished copy). Not for the positioning canvas itself — use positioning-mapper; not for finished blog posts or pages — use content-writer; not for ad or email units — use each discipline creative builder; not for claim adjudication — use offer-claims-registry. 消息屋/PR-FAQ/价值支柱/发布叙事
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。
Use when the user asks to "author our durable brand message hierarchy", "build the brand message house that seeds the canon", or "define the main narrative, three pillars, and tagline for the whole brand"; produces the DURABLE brand message system — main narrative, three value pillars, per-persona proof points (each labeled Measured / User-provided / '[needs source]'), and the tagline/one-liner — that seeds the narrative-registry canon and from which every per-launch house derives. Not for a single launch's message house or PR-FAQ — use message-house-builder; not for claim adjudication — use offer-claims-registry; not for the change-narrative arc itself — use strategic-narrative-designer. 品牌消息屋/主叙事/三支柱/标语/持久叙事 canon
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。
Use when the user asks to "measure how our narrative is landing", "track echo rate against our canon lexicon", or "check how AI answer engines describe our brand"; produces a resonance report — echo rate (overlap of market language with the narrative-registry canon lexicon, method declared), AI-answer perception via tavily.py --answer (proxy-labeled), share-of-voice on a locked competitor panel (reusing share-of-voice-tracker), and resonance signals from bluesky.py / gdelt.py / pageviews.py — every number labeled Measured / proxy / User-provided, feeding the TALE E dimension and the upstream of the E1 evidence-integrity veto. Not for rebuilding share-of-voice machinery — use share-of-voice-tracker; not for own-site GA4/GSC analytics — use performance-monitor; not for scoring TALE profile result — use narrative-quality-auditor; not for adjudicating claims — use offer-claims-registry. 回声率/AI回答感知/份额之声/共鸣信号
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。
Use when the user asks to "codify our brand voice", "define naming rules for our products and tiers", or "write the tone-of-voice guide with banned phrases"; produces the brand-level voice canon (register, tone spectrum, banned-phrase list, few-shot examples drawn only from the brand's own published material) and the naming tax (product / feature / tier naming rules plus approved and banned terms) that seeds the narrative-registry canon and that every channel's voice adaptation points up to. Not for per-platform voice adaptation — use channel-registry's voice-dossier; not for finished copy or blog posts — use content-writer; not for the message hierarchy itself — use message-system-architect; not for claim adjudication — use offer-claims-registry. 品牌语气/词汇表/命名税/禁用词/品牌语言规范
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。
Use when the user asks to "map what our buyers believe", "capture the objections we keep hearing", or "find the switching forces that move the beachhead"; produces a belief map of the beachhead — held beliefs and mental models, the recurring objections and their reframes, and the JTBD four forces (push of the problem, pull of the new, anxiety of switching, habit of the present) — each item sourced from interviews or win-loss notes (User-provided) and labeled Measured / User-provided / Estimated, with any unverified quote or comparative claim marked "[needs source]" and routed to the claims candidates, never adjudicated here. Not for demographic or persona profiling — use audience-mapper; not for the positioning canvas — use positioning-truth-tracer. 受众信念/异议地图/切换四力/流失语言
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。
Validate Claude Ads contracts, scoring inputs, run bundles, capabilities, source freshness, safety, installation, uninstall, or release readiness. Use for ads validate, ads status, ads next, stale claims with missing tool access, maturity checks, ownership-manifest uninstall, preserving unrelated ads-* skills, checksum verification, preflight, QA, or release audits. Missing access must demote stale evidence for the run and block dependent release-current claims before recovery guidance.
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。
リファクタリングを「更新されたチームの理解と、コードが表現する過去の理解の差分同期」と捉え、今回得た新知識（要求・ドメイン知識・制約）が naming と responsibility へ反映されているか、boundary が現在の understanding を表現しているか、diff の削除行やコメントに現れる過去の design history と constraint を失っていないかを diff-time で確認する。Knowledge Delta の signal は diff（追加/削除された hunk・コメント・ADR 参照）を一次情報とし、PR 本文が供給されるときは補助に用い、不確実なら question に留める。ドメイン用語の一貫性は ubiquitous-language-naming、集約/コンテキスト境界の設計判断は bounded-context-language、投機的抽象化・caller special-case は altitude-generalization、スコープ逸脱/前提破壊は fix-scope-integrity、振る舞い変更と構造変更の分離は behavior-structure-separation、完了主張の反証は refactor-claim-audit へ委譲する
検査器（linter / ガード / CI ゲート / バリデータ / フィルタ）そのものの検出力が差分で静かに落ちていないか、また新設した検査が実際に検出できることが実証されているかを diff-time で検出する。Check 1 unverified detection reduction（検査器の検出ロジック（正規表現・除外パターン・解決集合・allowlist・baseline）を変更して検出件数が減ったのに、減った 1 件ずつが誤検出であった根拠が示されないまま「誤検出を潰した」「N → M 件に減少」と改善として主張している）、Check 2 unproven guard detection power（新しい検査ロジック・CI ゲート・バリデーションを追加・変更したのに、検出すべき入力を注入して検出されること／検出すべきでない入力で黙ることを示す変異注入の証跡が無く、「追加した」「CI が緑」を検出力の証拠として扱っている）、の 2 Check を対象とする report-only。テスト自身のアサーションが実質何も検証していない構造は test-assertion-effectiveness、レビュー基準・品質ゲートの明示的な弱体化（ルールの削除・閾値の引き下げ・suppression entry 追加・lint ルールの無効化）は review-criteria-integrity、リファクタ完了主張一般の検証は refactor-claim-audit、workflow の permissions と action pin は gha-workflow-security、設定ファイルの構文・型の妥当性は config-json へ委譲する
リファクタリング（構造変更）と機能変更（behavior change）が同一 diff に混在していないかを識別し、structural change に対する external behavior preservation の証拠（テスト・型・静的解析）が揃っているかを diff-time で確認する。証拠不足のときは Characterization Test 追加を促すか question とする。DTO/公開 API の破壊的変更検出は api-compatibility、完了主張の反証や抽出リファクタの性能特性退行は refactor-claim-audit、新知識の命名・責務・境界への反映は knowledge-to-code-alignment、投機的抽象化・スコープ逸脱は altitude-generalization / fix-scope-integrity へ委譲する
Deep, iterative internet research for current facts, source triangulation, contradiction checks, due diligence, competitive analysis, claim verification, "latest" lookups, and cited reports. Use when the user asks to browse, search, research the web, verify claims, compare sources, gather primary sources, investigate companies, people, products, technical topics, news, events, markets, or create AKBP-ready durable research evidence.
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。
Use when about to claim work is complete, fixed, or passing, before committing or creating PRs - requires running verification commands and confirming output before making any success claims; evidence before assertions always
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。