AI / LLMs
TronLink's developer documentation is published in machine-readable form so AI agents and LLM-based tools can discover, read, and integrate it directly.
Allowed uses vs. crawler policy
This site is curated for retrieval / RAG / inference-time grounding — the llms.txt and llms-full.txt bundles below are the canonical ingestion points for that case, and they are explicitly maintained for AI use.
Training-time bulk crawling is a separate question and is governed by the site's robots.txt plus the Content-Signal directives served at the edge. Some training-only user-agents are currently disallowed there; if you need a different policy for your use case, contact us rather than parsing the bundles to circumvent the signal.
Available endpoints
URL ≠ source filename. Source files carry
.en/.zhsuffixes (llms.zh.txt,llms-full.en.txt,llms-full.zh.txt) so the mkdocs i18n plugin can route them. In the deployed site the suffix is stripped — Chinese variants live under/zh/. Use the deployed URLs below; the source-name URLs (e.g./llms-full.en.txt) return 404.
| Endpoint (deployed URL) | Description |
|---|---|
| /llms.txt | Curated English index of key pages (llmstxt.org format) |
| /zh/llms.txt | Curated Chinese index — same layout, links into /zh/ pages |
| /llms-full.txt | Every English page concatenated for single-fetch ingestion (built from docs/llms-full.en.txt) |
| /zh/llms-full.txt | Every Chinese page concatenated for single-fetch ingestion (built from docs/llms-full.zh.txt) |
Production URLs: https://docs.tronlink.org/llms.txt, https://docs.tronlink.org/zh/llms.txt, and the matching llms-full.txt bundles under each locale root.
Which file should I use?
| Use case | URL |
|---|---|
| Navigate / find the right English page | /llms.txt — short, link-only map |
| Navigate / find the right Chinese page | /zh/llms.txt — same layout, Chinese descriptions |
| Ingest the whole English documentation in one request | /llms-full.txt |
| Ingest the whole Chinese documentation in one request | /zh/llms-full.txt |
Start with the index for your language and follow its links; fetch a full bundle when you need everything at once. The Chinese index points at /zh/ slugs; the English index points at root-level slugs.
Add to your AI tool
- Cursor — Settings → Features → Docs → Add new doc, paste
https://docs.tronlink.org/llms.txt. - Claude / MCP-capable agents — TronLink ships MCP servers for live wallet and chain access; see MCP Server TronLink and MCP TronLink Signer for host configuration. For documentation context, point the tool at the
llms.txtURL above. - Other tools — any tool that accepts a custom documentation URL can use the
llms.txt/llms-full.txtendpoints.
Coverage
The bundle covers the full documentation, including the AI/agent tooling:
- MCP Server TronLink — on-chain, multi-sig, and GasFree tools (Playwright + Direct API)
- TronLink MCP Core — framework library: schemas, tool definitions, flow recipes
- TronLink Skills — read-only agent skill set (wallet, token, market, swap, resource, staking)
- MCP TronLink Signer — MCP server wrapping the signer SDK
- TronLink Signer — standalone signer SDK
- TronLink CLI — command-line tool
Plus the DApp integration, mobile (DeepLink), and Reference (networks, glossary, FAQ) sections.
Notes for agents
- Start from
llms.txt— it is the authoritative map. Don't enumerate the site blindly. - For tool calls, branch on the structured
error.code/error.retryable, never on the human-readablemessage. - Read operations are safe to retry; signing / Remote Write operations require user approval (HITL) and must not be auto-retried — see each tool's Safety section.
- Default to testnets (
nile/shasta) when experimenting; usemainnetonly for real funds.