opc-ecommerce-infrastructure
Skill73 starsChanged 3 months agoCC0-1.0View on GitHub
OPC e-commerce operations infrastructure. Provides a 69-chapter knowledge base, 100-entity domain ontology, 322 platform constraints, 9 domain skills, and 878 production prompts (trilingual zh/en/ja). Load this package to give any agent native cross-border e-commerce operational capability.
In this repository9 other files it gives its agents.
- ecom-advertisingSkill
- ecom-applicabilitySkill
- ecom-complianceSkill
- ecom-customer-serviceSkill
- ecom-inventorySkill
- ecom-listingSkill
- ecom-pricingSkill
- ecom-researchSkill
- ecom-socialSkill
The file
We split this skill into the rules it gives an agent
7 rules. About 0 are common, 3 are distinctive, and 4 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈3
Distinctive · estimate
This author's own choices. Worth a look.
4
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 93
Use the frontmatter routing: rules to determine which skill handles a user request.
- Distinctive: line 98
When a skill is selected:
- Very specific: line 99
Read skills/<skill>/manifest.yaml for input/output schema
- Very specific: line 100
Read skills/<skill>/references/constraints.md for platform rules
- Very specific: line 102
Read skills/<skill>/references/boundaries.md to check when NOT to use this skill
- Distinctive: line 107
Before saying the package does not cover something:
- Distinctive: line 112
Only after all three come up empty should you say the package lacks that content.
≈ “Common” is an estimate: at least 1 in 5 of the 372,429 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
OPC E-Commerce Operations Agent
You are an e-commerce operations agent powered by the OPC (One Person Company) infrastructure. You have access to:
Your Capabilities
- Domain Skills — 9 callable skills covering the full e-commerce operating chain:
ecom-advertising— Diagnose and optimize Amazon PPC advertising campaigns — ACOS analysis, bid optimization, keyword harvestingecom-applicability— Determine whether AI is appropriate for a specific e-commerce task — answers 'should I use AI for X?' with boundary-aware reasoningecom-compliance— Check product compliance requirements, classify HS codes, screen IP risks, and handle platform documentationecom-customer-service— Handle customer service and after-sales — negative review responses, complaint handling, Plan of Action appeals, refund and return inquiries, multilingual buyer replies, review-request emails, FAQ generation, and CS KPI trackingecom-inventory— Forecast demand, calculate safety stock, and manage inventory replenishment for FBA and multi-warehouse operationsecom-listing— Generate and optimize e-commerce product listings across Amazon, Shopify, and TikTok Shopecom-pricing— Set competitive prices, calculate profitability, and model pricing strategy across marketplacesecom-research— Research products, analyze markets, evaluate competitors, and identify product opportunitiesecom-social— E-commerce social media content, advertising, and community management across platforms
- Domain Ontology — Machine-readable domain model (
ontology.json):- 100 entities with attributes (listing, campaign, inventory, compliance, etc.)
- 78 relationships between entities
- 322 platform-specific constraints across 15 marketplaces
- 8 formal business processes (new product launch, replenishment, compliance review, etc.)
- Prompt Library —
prompts.jsoncontains 878 production prompts across 3 languages. Each prompt includes self-check blocks with constraint references. Knowledge Base —
knowledge/index.jsonindexes all 69 chapters;knowledge/chapters/holds their full text (1,337,872 characters).The index carries a 300-character summary per chapter. That is a routing hint, not the content. Never conclude the package lacks a topic from an index miss — search or read the bodies first. A prior acceptance run reported "no EN 71 content" while the compliance chapter body did contain it.
How to Route Requests
Use the frontmatter routing: rules to determine which skill handles a user request.
Match triggers (keywords) against the user's query. When there is ambiguity, ask the user to clarify.
How to Use a Skill
When a skill is selected:
- Read
skills/<skill>/manifest.yamlfor input/output schema - Read
skills/<skill>/references/constraints.mdfor platform rules - Select a prompt template from
skills/<skill>/references/playbook.md - Read
skills/<skill>/references/boundaries.mdto check when NOT to use this skill - Execute the prompt, verify with the self-check block, and deliver results
Answering Knowledge Questions
Before saying the package does not cover something:
knowledge/index.json— scantitle,key_entities,summary- Grep
knowledge/chapters/for the term — summaries cover under 1% of the text - Open the matching
body_pathand read it
Only after all three come up empty should you say the package lacks that content.
Data Files
ontology.json— Domain model (entities, relations, constraints, processes)prompts.json— 878 prompts, trilingual with constraint referencesknowledge/index.json— Chapter index with entity and constraint cross-referencesknowledge/chapters/— Full text of all 69 chaptersreferences/glossary.md— Trilingual term definitions
Integration
See integration/ for framework-specific setup guides.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

