AI / LLMs¶
JustLend DAO provides machine-readable documentation endpoints optimized for LLMs and AI coding tools. Use these resources to give your AI assistant accurate, up-to-date context about the JustLend protocol.
Available endpoints¶
| File | Description |
|---|---|
llms.txt |
Curated index of key pages with one-line descriptions. Start here for quick lookups and scoped questions. |
llms-full.txt |
Condensed full-text snapshot of the documentation for single-fetch LLM consumption. Use when your tool supports larger context. |
documents/aidocs |
Compact AI/RAG-oriented pages with source-of-truth rules, tool mappings, safety policy, Chinese/English FAQs, and MCP tool catalog. |
Machine-readable sources
For API and contract integrations, also prefer the OpenAPI spec, contract directory, contract directory JSON Schema, and JSON ABI files under /developers/abis/.
Which file should I use?¶
| Use case | Recommended |
|---|---|
| Quick lookups, scoped questions, page discovery | llms.txt |
| Full protocol understanding, complex integrations, offline ingestion | llms-full.txt |
| RAG chunks, benchmark questions, Chinese/English user intent routing | documents/aidocs |
| Public HTTP API calls and response schemas | justlend_apis.yaml |
| Contract addresses in Base58 / EVM hex / TRON hex | contracts.json |
Validating a copy of contracts.json (schema check) |
contracts.schema.json (JSON Schema 2020-12) |
| Contract calls, event decoding, and agent tooling | developers/abis/ |
| MCP tool routing and safety annotations | MCP Tool Catalog |
AI Docs scoring and retrieval hints¶
The documents/aidocs section is intentionally more repetitive and task-oriented than the human docs. It is designed for RAG scoring, benchmark questions, and agent routing:
- each page has frontmatter
title,description, andtags; - each page answers one narrow intent such as market data, account position, lending action, sTRX staking, or energy rental;
- safety-sensitive workflows link to the MCP safety policy before any write action;
- Chinese and English common questions are mapped to the same source-of-truth tools.
Notes for agents¶
- Treat the OpenAPI spec, MCP tools,
contracts.json, and ABI JSON files as machine-readable sources of truth. - Treat rendered pages as human-readable explanations and examples.
- Respect documented precision rules: jTokens use 8 decimals; underlying assets use their own decimals; rates and mantissas are scaled by
1e18. - Use Nile testnet for integration testing and Mainnet only when users explicitly intend production transactions.