Skills worth trusting.
Agent skills from Anthropic, OpenAI, skills.sh, ClawHub, SkillsMP and other reputable sources. Savant assesses them for security risks and evaluates model-generated test cases using independent scoring. Workspaces import any skill into their own repositories through a reviewed pull request. Every listing on each hub is enumerated; packages are fetched, scanned and evaluated most-popular first as daily capacity allows.
tdd-guide
Test-driven development skill for writing unit tests, generating test fixtures and mocks, analyzing coverage gaps, and guiding red-green-refactor workflows across Jest, Pytest, JUnit, Vitest, and Mocha. Use when the user asks to write tests, improve test coverage, practice TDD, generate mocks or stubs, or mentions testing frameworks like Jest, pytest, or JUnit.
pricing-packaging-planner
Use when the user asks to "plan launch pricing", "design pricing tiers / packaging", or "set up a launch discount / early-bird offer"; produces a launch pricing and packaging plan — tier structure and naming, a value-to-price map aligned to the message house, a launch-offer ladder with a true deadline reason, beta / early-adopter pricing with a post-launch graduation path, and guarantee / refund terms — with every price claim and offer term submitted to memory/events/claims.ndjson. Not for the canonical offer / claim record — use offer-claims-registry; not for paid-newsletter subscription economics — use newsletter-monetization-planner; not for ad bidding — use bid-strategy-planner. 发布定价/梯度打包/早鸟优惠/保证设计
inbox-placement-monitor
Use when the user asks to "track where my emails are actually landing after I send", "read my seed-list inbox vs spam vs promotions results", "trend my Gmail Postmaster / Microsoft SNDS reputation", or "did placement drop after my last send"; produces a per-provider inbox/spam/promotions placement read, a domain/IP reputation trend from Postmaster + SNDS, a send-over-send delta with named regressions, and a reusable SEND-S placement snapshot on your own exported telemetry. Not for the pre-send SPF/DKIM/DMARC auth pre-flight — use deliverability-qa; not for computing the EQS or running the vetoes — use email-quality-auditor. 邮件收件箱落点监测/收件箱vs垃圾邮件/Postmaster声誉趋势
conversion-value-mapper
Use when the user asks to "set up conversion values so tROAS optimizes profit not orders", "map margin onto my purchase value", "build value rules for lead / phone / signup conversions", or "stop bidding to revenue when I care about profit"; defines and QAs the conversion VALUE model — per-conversion values, margin/net-value adjustment, static-vs-dynamic value rules, proxy values for non-revenue actions, and a value-vs-count sanity check — as a value-model spec plus a pre-launch value QA sheet. Not for whether the tag fires or UTMs are clean — use conversion-signal-qa; not for cross-platform double-count de-dup — use attribution-reconciler; not for scoring R1/R2 — that is a scored veto in ad-account-auditor. 付费广告转化价值建模/利润出价/价值规则QA
da-auth
Use this when another step needs to call the admin.da.live API, for example before pushing HTML content, listing documents, or triggering a DA preview, and you do not already have a valid DA_TOKEN in scope from an earlier step in the same session. Covers obtaining a valid Adobe IMS access token for the DA (Document Authoring) API.
launch-monitor
Use when the user asks to "monitor my launch", "track our Product Hunt / Hacker News ranking", or "watch the launch window"; runs the T-0 to T+30 window watch — pre-launch instrumentation verification (UTM/event checks, the upstream of RAMP P1), HN rank/points/comments polling with a comments-over-points flamewar early-warning (Estimated heuristic), PH votes/featured status, store charts and reviews, news echo, D0/W1/M1 KPI snapshots vs targets, spike-vs-sustain and owned-capture reads, and alert thresholds against the launch-tier KPI targets. Not for launch-day go/rollback calls — use launch-day-conductor; not for metric deep-dives — use performance-analyzer; not for SEO rank tracking — use rank-tracker. 发布监控/排名轮询/火焰战比/spike-sustain
consent-registry
Use when the user asks to "log this subscriber's opt-in", record unsubscribes/complaints, or query lawful basis; curates pseudonymous consent facts through the append-only consent stream and applies suppression/erasure tombstones immediately. Not for SEND scoring — use email-quality-auditor; not for building segments — use list-segment-builder. 订阅同意台账/实时退订抑制/合法性依据登记
reactivation-specialist
Use when the user asks to "build a win-back campaign", "re-engage lapsed subscribers", "run a re-permission / re-consent sweep", or "sunset my dead list"; produces a closed-loop reactivation program — a lapsed-cohort definition, a staged offer ladder, a re-consent (re-permission) capture step, and a sunset-confirm / suppression rule. Owns none of the SEND-N sub-item notes: engagement-decay / sunset is email-sequence-designer's and preference-center / frequency options is preference-frequency-manager's — this skill references those notes, it does not re-emit them. Not for the general lifecycle flows (welcome/cart/post-purchase) — use email-sequence-designer; not for the preference-center / opt-down ladder — use preference-frequency-manager; not for the consent record itself — use consent-registry; not for computing EQS or the N1 veto — use email-quality-auditor. 流失召回/重新授权/沉默用户清理
use-arc
Provide instructions on how to build with Arc, Circle's blockchain where USDC is the native gas token. Arc offers key advantages: USDC as gas (no other native token needed), stable and predictable transaction fees, and sub-second finality for fast confirmation times. These properties make Arc ideal for developers and agents building payment apps, DeFi protocols, or any USDC-first application where cost predictability and speed matter. Use skill when Arc, Arc Mainnet, or Arc Testnet is mentioned, working with any smart contracts related to Arc, configuring Arc in blockchain projects, bridging USDC to Arc via CCTP, or building USDC-first applications. Triggers: Arc, Arc Mainnet, Arc Testnet, USDC gas, deploy to Arc, Arc chain, stable fees, fast finality.
observability-driven-testing
Use production telemetry as INPUT to design new tests. Covers OpenTelemetry integration with tests, trace-based assertions, log-informed test creation, production-error analysis for coverage gaps, and telemetry-driven test prioritization. Use when: "trace-based testing," "design tests from logs," "OpenTelemetry assertions," "production errors point to test gaps," "telemetry-driven testing." Not for: safe rollout techniques (flags, canary) during release — use testing-in-production. Not for: scheduled post-deploy probes — use synthetic-monitoring. Not for: triaging CI failures — use ai-bug-triage. Related: testing-in-production, synthetic-monitoring, qa-metrics, ai-bug-triage.
auth-tool-cloudbase
CloudBase auth provider configuration and login-readiness guide. This skill should be used when users need to inspect, enable, disable, or configure auth providers, publishable-key prerequisites, login methods, SMS/email sender setup, or other provider-side readiness before implementing a client or backend auth flow.
kalshi
Kalshi prediction markets — events, series, markets, trades, and candlestick data. Public API, no auth required for reads. US-regulated exchange (CFTC). Covers football (EPL, UCL, La Liga), basketball, baseball, tennis, NFL, hockey event contracts. Use when: user asks about Kalshi-specific markets, event contracts, CFTC-regulated prediction markets, or candlestick/OHLC price history on sports outcomes. Don't use when: user asks about actual match results, scores, or statistics — use the sport-specific skill: football-data (soccer), nfl-data (NFL), nba-data (NBA), wnba-data (WNBA), nhl-data (NHL), mlb-data (MLB), tennis-data (tennis), golf-data (golf), cricket-data (cricket), cfb-data (college football), cbb-data (college basketball), or fastf1 (F1). Don't use for general "who will win" questions unless Kalshi is specifically mentioned — try polymarket first (broader sports coverage). Don't use for news — use sports-news instead.
optimized-nextjs-typescript
Optimized Next.js TypeScript best practices with modern UI/UX, focusing on performance, security, and clean architecture
engagement-inbox-manager
Use when the user asks to "triage our comments, DMs, and mentions", "draft replies to this thread", "can we repost this fan post", or "set up inbox SLAs and an escalation path"; produces a ranked triage queue with register detection (sincere / ironic / performative / parasocial, sentiment-inversion table included — "this is so bad" under a comedy register is praise), a commenter taxonomy (troll monitor-only / rager / misguided / unhappy-customer / advocate) with a response-tier ladder and per-channel SLAs, an escalation matrix ending at the crisis path, a moderation ladder plus house rules for owned spaces, and a UGC curation-and-rights mode whose dated permission entries route to the channel registry — every reply is a ranked DRAFT a human posts; nothing is ever auto-sent. Not for launch-window feedback triage — use launch-feedback-synthesizer. 评论私信提及分诊/语域识别/回复草稿/UGC授权
social-pulse-monitor
Use when the user asks to "monitor brand mentions", "set up social listening", "did anything spike about us this week", or "watch these accounts for buying triggers"; runs always-on keyless listening — a versioned listening-query architecture (brand variants incl. misspellings and 中文 names, exclusion terms, per-source syntax for HN Algolia / Bluesky / GDELT / Tavily), a mention sweep with six-class triage (crisis / bug / lead / praise / question / spam) each routed with an SLA, a 7-day rolling baseline with spike flags (the crisis trigger input), a B2B trigger watchlist (funding / hiring / launch signals), and an explicit coverage disclosure where X/IG/TikTok/LinkedIn/小红书 numbers are proxy-labeled, never Measured. Not for launch-window telemetry (T-0→T+30 rank and review polling) — use launch-monitor. 舆情监听/品牌提及监测/基线突刺/触发信号
ax
Use the ax CLI instead of curl + throwaway parsing scripts whenever you fetch a URL, explore an unknown web page, or extract structured data from HTML. Trigger whenever you are about to write an inline script (python3 heredoc, node -e, regex over HTML) or a bare curl for one-off web fetching, scraping, or page exploration.
phoenix-cli
Debug LLM applications using the Phoenix CLI. Fetch traces, analyze errors, structure trace review with open coding and axial coding, inspect datasets, review experiments, query annotation configs, and use the GraphQL API. Use whenever the user is analyzing traces or spans, investigating LLM/agent failures, deciding what to do after instrumenting an app, building failure taxonomies, choosing what evals to write, or asking "what's going wrong", "what kinds of mistakes", or "where do I focus" — even without naming a technique.
omni-usage-logs
Access detailed call logs and usage analytics. Filter by provider, model, time range, status, and cost. Export logs and aggregate token usage across all connections.
amazon-location-service
Integrates Amazon Location Service APIs for AWS applications. Use this skill when users want to add maps (interactive MapLibre or static images); geocode addresses to coordinates or reverse geocode coordinates to addresses; calculate routes, travel times, or service areas; find places and businesses through text search, nearby search, or autocomplete suggestions; retrieve detailed place information including hours, contacts, and addresses; monitor geographical boundaries with geofences; or track device locations. Covers authentication, SDK integration, and all Amazon Location Service capabilities.
positioning-truth-tracer
Use when the user asks to "check our positioning against what we can actually ship", "trace which differentiators we can defend", or "reconcile the positioning canvas with the claims ledger"; reconciles the reused positioning canvas against the shippable stage and the claims ledger to produce a differentiation truth set — every differentiating claim verifiable or marked '[needs source]' — that TALE-T1 is judged against. Not for building the canvas — use positioning-mapper; not for adjudicating claims — use offer-claims-registry; not for authoring the message house — use message-system-architect. 定位真相/差异化校准/可交付现实/主张核对
list-hygiene-monitor
Use when the user asks to "watch my list health over time", "flag decaying / unengaged subscribers on a schedule", "why is my open rate drifting down / bounces creeping up", or "build me a re-permission and prune worklist"; runs the scheduled SEND list-decay + suppression-drift watch — an engagement-recency cohort read (30/90/180/365-day), hard-bounce and spam-complaint trend vs benchmark, suppression-list growth/leakage check, and a segmented re-permission / sunset / prune worklist tied to SEND S (list hygiene) and E (engagement-decay) sub-items. Not for the one-time pre-send authentication pre-flight — use deliverability-qa; not for the consent/suppression record itself — use consent-registry; not for computing the EQS or enforcing vetoes — use email-quality-auditor. 邮件列表健康度监控/退订漂移/沉睡用户清理
early-access-designer
Use when the user asks to "design an early access program", "set up a waitlist and beta stages", or "define beta graduation criteria"; produces a waitlist→concept→alpha→beta→GA stage ladder with per-stage purpose and opt-in semantics, quantified graduation criteria per stage (labeled Estimated), a cohort-gating and invite-throttling plan, tester recruitment with launch-day social-proof prep, a feedback-loop spec where every status change notifies its subscribers, and a referral-loop mechanism spec (invite codes, anti-abuse). Not for waitlist acquisition strategy or the capture-flow spec — use list-growth-designer; not for the canonical stage record — use launch-registry. waitlist/内测阶梯/抢先体验/毕业标准/反馈闭环
felo-slides
Generate PPT/slides from prompts, local images, PDFs, and other files with the Felo PPT Task API. Use when users ask to create, make, generate, or export presentations or slide decks; turn an attached/local file into slides; include an original image in a PPT; or explicitly invoke /felo-slides. Handles file upload, API key checks, task creation, polling, and final ppt_url output.
senior-prompt-engineer
World-class prompt engineering skill for LLM optimization, prompt patterns, structured outputs, and AI product development. Expertise in Claude, GPT-4, prompt design patterns, few-shot learning, chain-of-thought, and AI evaluation. Includes RAG optimization, agent design, and LLM system architecture. Use when building AI products, optimizing LLM performance, designing agentic systems, or implementing advanced prompting techniques.
analyzing-mlflow-trace
Analyzes a single MLflow trace to answer a user query about it. Use when the user provides a trace ID and asks to debug, investigate, find issues, root-cause errors, understand behavior, or analyze quality. Triggers on "analyze this trace", "what went wrong with this trace", "debug trace", "investigate trace", "why did this trace fail", "root cause this trace".
launch-feedback-synthesizer
Use when the user asks to "triage launch feedback", "cluster reviews, comments, and board posts into themes", or "set up a you asked, we shipped loop"; produces a feedback theme digest (frequency, severity, representative quotes per theme), an open→planned→started→completed/declined status loop with duplicate-merge and notification rules, shipped-change announcement material, and a compliant social-proof harvest protocol (never incentivized store reviews). Not for repurposing or amplifying the harvested proof — use content-amplifier; not for executing testimonial outreach threads — use outreach-manager. 反馈分诊/状态环/社证收割/评测合规
newsletter-monetization-planner
Use when the user asks to "monetize my newsletter", "build a sponsorship rate card", or "model paid-subscription revenue"; produces a revenue model (paid tiers, ad/sponsorship inventory + CPM/flat rate card, referral/boost loops), a list-growth ↔ revenue projection, and honest-offer / disclosure checks for the SEND-D lever. Not for scoring the whole program or running D1 — use email-quality-auditor; not for the return math — use roi-calculator; not for the post-click page — use landing-optimizer. 邮件newsletter变现/赞助刊例/付费订阅测算
cloud-storage-web
Complete guide for CloudBase cloud storage using Web SDK (@cloudbase/js-sdk) - upload, download, temporary URLs, file management, and best practices.
pexels-video-downloader
Download free Pexels videos in 4K quality with batch support
creative-director
AI creative director with recursive self-assessment. Generates concepts using world-class methodologies (SIT, TRIZ, Lateral Thinking, bisociation), scores against 6 weighted criteria with Cannes/D&AD/HumanKind calibration, and recursively refines until the 9+ threshold is reached. Accepts briefs in any format — text, voice transcript, PDF, or raw notes. Use when the user asks to generate creative concepts, brainstorm campaign ideas, develop a Big Idea or campaign platform, evaluate or critique existing creative work, find consumer insights, or shares a brief for ideation — including activations, PR-stunts, brand utility, experiential, and non-advertising ideas. Calibrates against a library of 569 legendary campaigns (P01-P18 pattern map) to detect saturation and ensure originality. Do not use for media planning, production budgeting, brand identity/logo design, copywriting final drafts, or market research data collection.
social-measurement-loop
Use when the user asks to "run the weekly social readout", "which denominator does our engagement rate use", or "which posts won this week and what changes next cycle"; produces the organic-social metric dictionary (every rate names its denominator — ERR engagement-by-reach vs ERI by-impressions vs ER-by-follower — locked across periods), median-not-mean per-post rollups with organic and boosted separated, EMV as labeled exec-translation only (never inside any score), an attributed CHAOSS/Orbit-style community-health readout with employees excluded, and the best/worst-performer write-back the next calendar cycle consumes. Not for dollar-ROI math or the ECHO profile result gate verdict — use roi-calculator and social-quality-auditor. 社媒周报/互动率分母/指标字典/复盘回写
checking-freshness
Quick data freshness check. Use when the user asks if data is up to date, when a table was last updated, if data is stale, or needs to verify data currency before using it.
launch-day-conductor
Use when the user asks to "run my launch day", "build a launch day runbook / war room", or "decide CONTINUE or ROLLBACK after the push"; produces a pre-conditions gate check (launch-readiness-auditor SHIP verdict + the authoritative date in launch-registry — missing either stops the skill), a dated hour-blocked runbook with owners (morning irreversible pushes, daytime monitoring loop, evening consolidation), a forced observation-window verdict after every irreversible action against pre-declared kill criteria, a P0-P3 incident ladder with rollback playbooks, and T-0 status lines for the registry proposal protocol. Not for channel submission content and platform rules — use community-launch-runner; not for media replies — use press-media-relations. 发布日runbook/作战室/观察窗/回滚裁决/发布日指挥
narrative-baseline-mapper
Use when the user asks to "map what our surfaces say today", "inventory our current messaging", or "find the gap between what we say and what we mean"; produces the narrative baseline — a surface-by-surface inventory of what every owned touchpoint (homepage, pricing, docs, decks, social bios, email footers) claims RIGHT NOW, each line labeled Measured / User-provided / Estimated, plus a per-surface gap read vs the intended message and the drift-baseline snapshot the Evaluate phase measures future drift against. Not for authoring the canon — use message-system-architect; not for scoring the surfaces or running the vetoes — use narrative-quality-auditor. 现状叙事盘点/各触点口径/意图差距/漂移基线
arize-instrumentation-health
Audits instrumentation health of existing Arize traces. Runs deterministic checks over a bounded span sample (orphaned/uncategorized/duplicate spans, flat structure, blank root I/O, unset status, missing token counts or children) and returns a ranked report. Use when the user asks why traces look empty/flat/broken, wants to verify instrumentation is healthy, find instrumentation issues, or why evals or token/cost dashboards show n/a or zero. To debug app behavior or errors, use arize-trace.
share-of-voice-tracker
Use when the user asks to "track our share of voice", "what share of the conversation do we own vs competitors", or "trend our SOV this quarter"; computes SOV% = brand mentions ÷ (brand + competitor panel mentions) per platform per period on a LOCKED competitor panel — a panel switch invalidates the trend (restart the series and log the break; the ECHO O3 rule) — plus a sentiment-weighted SOV variant (sentiment labeled Estimated unless human-coded) and a Wikipedia-pageviews attention-share alternative; built from keyless listening connectors, gdelt.py news echo, and user exports — public counts only, closed platforms are never scraped. Not for backlink or offsite SEO signals — use offsite-signal-analyzer. 声量份额/竞品声量对比/提及份额/注意力份额
tailscale
Guide for installing, configuring, and managing Tailscale, headscale, and the Tailscale product family. Covers core mesh VPN (exit nodes, subnet routers, access controls, SSH, MagicDNS), Docker and Kubernetes integration, CI/CD pipelines, ephemeral nodes, infrastructure access, site-to-site networking, app connectors, Aperture (AI/LLM gateway for governance, cost control, usage visibility), device posture, MDM, SCIM provisioning, SSH and kubectl session recording, tsrecorder, Taildrop, Tailscale Serve, Funnel, and building Go applications that embed Tailscale via the tsnet library. Use when someone asks about Tailscale networking, mesh VPN, VPN replacement, containers, Kubernetes operator, CI/CD runners, device management, session recording, audit logging, LLM API access, AI cost control, file sharing, exposing internal services, or writing a Go program that joins a tailnet as its own device — even when they describe the scenario without naming the product.
ralph-tui-create-beads
Convert PRDs to beads for ralph-tui execution. Creates an epic with child beads for each user story. Use when you have a PRD and want to use ralph-tui with beads as the task source. Triggers on: create beads, convert prd to beads, beads for ralph, ralph beads.
launch-window-planner
Use when the user asks to "pick a launch date", "plan the launch window", or "set the embargo and lift time"; produces a candidate-window comparison table (conflict / tailwind / risk per window) built from industry-event cycles and the competitor launch calendar, a launch-week vs rolling-release format call, store-review buffer padding (labeled Estimated), and an embargo window definition (lift moment + timezone) submitted to the launch registry as a candidate. Not for judging the cultural moment itself — use trend-spotter; not for launch-day execution — use launch-day-conductor. 发布择时/发布窗口/竞品日历/禁运期窗口/审核缓冲
launch-retro-analyzer
Use when the user asks to "run a launch retro / post-mortem", "compare launch results vs targets by channel", or "decide what to keep or kill for the next launch"; produces a structured D1/W1/M1 retrospective — a per-channel actual-vs-target table (UTM-attributed own analytics as the truth column, platform self-reported numbers as reference, every figure labeled Measured / User-provided / Estimated), a 5-Whys chain on the single largest miss, keep / kill / change decisions per channel, 3-5 actionable learnings for the next launch, and an outcome snapshot submitted to the launch registry. Not for return math (CPA / ROI) — use roi-calculator; not for the stakeholder-facing report writeup — use report-generator; not for a metric deep-dive — use performance-analyzer. 发布复盘/渠道归因/5-Whys/keep-kill
message-test-designer
Use when the user asks to "test our messaging before we scale it", "design a message-market-fit panel", or "run a 5-second comprehension test on our new tagline"; produces a message-test design spec — hypothesis, panel and recruit criteria, comprehension / 5-second / message-market-fit (Wynter-style) protocols, stimulus set drawn from the canon, success thresholds, and a stop/revise decision rule — for the TALE Evaluate phase so the message is validated before any paid scale. It designs the test; it never runs the experiment or adjudicates a claim. Not for running the panel or A/B experiment — use send-experiment-designer or ad-test-designer; not for analyzing the results — use performance-analyzer; not for authoring the message itself — use message-system-architect. 消息测试/理解度测试/面板设计/五秒测试/消息市场契合
narrative-cascade-planner
Use when the user asks to "plan how our narrative lands on every surface", "write per-surface message-match specs", or "brief each creative builder from the canon"; maps the narrative-registry canon onto every flagship surface (homepage, pricing, store listing, sales deck, social bio, docs, email) as a per-surface message-match spec — what claim, pillar, proof, and objection reframe each surface must carry — and cuts a handoff brief to each discipline's creative builder (content-writer, ad-creative-builder, email-creative-builder, social-creative-builder, message-house-builder), so the same story lands consistently everywhere. Not for writing the finished surface copy — use each discipline creative builder; not for the tier-scoped launch asset manifest — use launch-asset-packager; not for scoring cross-surface consistency — use narrative-quality-auditor; not for claim adjudication — use offer-claims-registry. 叙事落地/逐面消息一致/创意交接简报/画布传导
cmux-cua
Use only after the user explicitly asks for cmux Computer Use through the cmux-cua skill: drive real macOS apps from a cmux agent session via the bundled engine (accessibility tree + screenshots, click/type/scroll/drag, branded cursor), or explain its user-directed permission setup. Reading or discovering this skill is not consent.
design-loop
Autonomous multi-page site builder using a baton-passing loop. Each iteration reads a task from .design/next-prompt.md, generates a page in HTML/Tailwind, integrates it into the site, verifies visually, then writes the next task to keep the loop alive. Use whenever the user asks to build an entire site autonomously, build all pages of a site, generate multiple pages in sequence, or run a 'design loop' / 'baton loop' / 'autonomous site build' — even if they say 'just keep going' or 'build the next page' or 'next page' mid-flow.
voice-dossier-builder
Use when the user asks to "codify how our founder sounds", "build a founder voice dossier", or "set our content pillars"; runs an 80%-extraction interview over the user's OWN posts, emails, and decks (never competitor scraping, never an invented persona) and produces the versioned voice record — a per-platform register map (incl. 小红书/微信公众号), banned phrases, per-context disclosure lines (the ECHO C2 upstream), a few-shot bank built exclusively from own posts, and 3-5 content pillars with Estimated %-allocations — submitted via memory/events/channels.ndjson for channel-registry to store as voice-dossier.md, the record every Craft-phase skill reads first. Not for the brand-level voice canon — use brand-language-codifier; not for audience/persona research — use audience-mapper. 声音档案/品牌语气/创始人语气/内容支柱/披露声明
platform-norm-profiler
Use when the user asks to "build the norm card for this platform", "what are the char limits and visible-fold cutoffs here", "is the LinkedIn link-in-first-comment thing documented or folklore", or "which of our platform cards are stale"; maintains the dated, versioned per-platform norm cards in the references/platforms/ pack — char limits, visible-fold cutoffs, hashtag norms, format/aspect specs, link and first-comment placement, disclosure-label mechanics, algorithm emphases (e.g. 小红书 search+saves weighting) — every row labeled platform-documented (official doc, Measured) or Estimated-folklore (named source) with last-verified and review-by dates, and any card past its review date flagged stale rather than trusted. Extends the single pack in place; never forks a second one. Not for picking which channels to run — use channel-portfolio-planner. 平台规范卡/字符限制/折叠线/话题标签/算法侧重/过期标记
paid-measurement-loop
Use when the user asks to "read back" a paid campaign change, "did this ad change work", or "compare ROAS/CPA before and after"; reads ROAS/CPA against a control over a fixed readback window and returns a Promote / Keep-testing / Rollback / Unproven readback decision with the math delegated to roi-calculator. Not for RQS scoring or veto adjudication — use ad-account-auditor; not for the ROI ratio math — use roi-calculator; not for cross-channel rollups — use performance-analyzer. 付费广告复盘/ROAS回看/投放效果归因
clean-architecture
Guidelines for implementing Clean Architecture patterns in Flutter and Go applications, with emphasis on separation of concerns, dependency rules, and testability.