| name | chainlink-confidential-ai-attester-skill |
| description | Chainlink Confidential AI Attester: submit private documents to an LLM inside an AWS Nitro Enclave and get back a cryptographically attested result โ raw documents never leave the TEE. Use for these hackathon scenarios: (1) undercollateralized DeFi lending โ upload a bank statement, get an attested approved/denied JSON decision without exposing financials on-chain; (2) accredited investor verification โ check SEC Rule 501 qualification from brokerage statements privately; (3) KYC/AML screening โ analyse ID docs and transaction history inside a TEE, return a pass/fail with flags; (4) proof of reserves โ verify custodian balance reports against claimed reserves; (5) any use case where an AI must read sensitive user documents and the result needs a cryptographic proof of what model ran on what data. Trigger on: private inference, attested AI, TEE inference, confidential AI, or undercollateralized lending / KYC / accredited investor mentioned alongside document analysis. |
| license | MIT |
| compatibility | Designed for AI agents that implement https://agentskills.io/specification, including Claude Code, Cursor Composer, and Codex-style workflows. |
| allowed-tools | Read WebFetch Write Edit Bash |
| metadata | {"version":"0.0.1"} |
Chainlink Confidential AI Attester
Runs LLM inference inside Trusted Execution Environment (TEE). Documents go in, LLM analysis comes out โ the raw documents are never stored or exposed.
Beta product for the EthGlobal NYC hackathon. Get an API key at the Chainlink booth or via the #partner-chainlink channel in the EthGlobal Discord.
Playground UI: https://confidential-ai-dev-preview.cldev.cloud/playground โ easiest way to try it. Everything there maps 1:1 to the API calls below.
Workflow 1 โ Submit: POST /v1/inference
Auth: Authorization: Bearer $API_KEY โ always use an env var, never hardcode.
Request shape:
{
"model": "gemma4",
"system_prompt": "",
"prompt": "...",
"resources": [{ "filename": "doc.pdf", "content_type": "application/pdf", "content_base64": "<base64>" }],
"cre_callback": { "url": "https://..." }
}
cre_callback is optional โ omit it and poll instead.- Models:
gemma4 (images/general, default), qwen3.6 (long text).
- Prefer PNG over PDF for demos โ PDF preprocessing can take up to 5 minutes.
Response: 202 Accepted โ { "id": "...", "status": "queued" } โ save the id.
For curl examples and multi-language snippets โ references/code-examples.md
For full request/response spec, error codes, resource types โ references/api-reference.md
Workflow 2 โ Poll: GET /v1/inference/{id}
Poll every 2โ5 s until status is completed or failed.
Key fields on completion: output (LLM text), usage, completed_at.
For error symptoms โ references/troubleshooting.md
Writing Prompts That Work
Always enforce JSON output with two layers:
- System prompt โ keep the default unless you have a specific reason to change it.
- User prompt โ binary question + exact JSON schema to return
For per-use-case prompt templates (lending, KYC, accredited investor, proof of reserves) โ references/prompts.md
