Installable, open-source marketing skills for AI agents helping global teams research, position, localize, launch, and learn in China.
Each skill owns one clear job, names the evidence it needs, and shows where local judgment must take over.
Start from a complete job, follow the handoffs, and keep the human decision points visible.
Open workflow paths →Establish the competitive evidence, choose a position, lock the integrated brief, prepare the announcement, and turn results into the next decision.
Turn customer language into a content thesis, rebuild one source for local channel jobs, then edit and package the WeChat article for release.
Align the event job, plan the promotion backwards, convert cross-team work into an SOP, and review the registration-to-follow-up funnel.
Map China-market competitors, platform-native capabilities, local substitutes, and evidence gaps before making a positioning or entry decision.
Synthesize Chinese interviews, reviews, sales notes, support records, and social comments without erasing platform context or overstating the sample.
Reframe a global product for a specific China buyer, category, alternative, value promise, and proof set without relying on translated global messaging.
Turn a China business objective, audience insight, platform constraints, proof, budget, and timing into an executable cross-channel campaign brief.
Turn approved product facts and positioning into credible Chinese B2B launch copy, executive messages, WeChat posts, and sales-ready language.
Diagnose, restructure, edit, and package Chinese WeChat Official Account articles for reader progression, credibility, and publication readiness.
Synthetic cases show inputs, execution traces, evidence limits, and human checkpoints.
Open field cases →A synthetic B2B AI team tests China market entry. The case shows how local products, platform-native capabilities, agencies, and workflow substitutes become separate evidence pools.
A synthetic global B2B launch draft is factually sound but reads like translated headquarters copy. The case rebuilds the Chinese reading path while preserving approved claims.
A synthetic China report campaign exceeds download targets but misses qualified-lead and follow-up goals. The case reconciles platform, CRM, agency, and execution evidence.
Every skill turns local context, evidence standards, deliverable structure, and boundaries into a reviewable agent workflow.
Define the China buyer, platform, language, region, and decision before producing work.
Date time-sensitive claims and separate official facts, inference, and unknowns.
Inputs, workflow, owner handoffs, and acceptance criteria are explicit.
Review every instruction, dependency, and limitation before installation.
Install the complete system or start with one workflow. Compatible with Codex, Claude Code, Cursor, and other Agent Skills hosts.
npx skills add fffffrost/market-skills