> ## Documentation Index > > Fetch the complete documentation index at: https://lithi.ai/llms.txt > > Use this file to discover all available pages before exploring further. --- title: Copy a Compute setup prompt description: >- Copy a versioned Lithi setup prompt into your frontier agent, review its safety boundaries, and choose a measured next step for customer savings or Mac earnings. route: /docs/compute/setup-prompts page_id: docs-compute-setup-prompts page_type: howTo content_layer: guided documentation surface_profile: guided_docs audience: Compute customers and Mac providers voice: D reader_question: How do I copy and review a Lithi setup prompt? primary_action: Choose what work to send to Compute source_locale: en-US source_status: APPROVED source_version: docs-compute-setup-prompts-en-us-2026-09-04-v1 last_updated: '2026-09-04' robots: index claim_registry_pinned: true claim_registry_resync: "npm run governed:tsx -- scripts/build-compute-public-claim-registry.ts --write" --- ## Choose a prompt and keep your key safe Use this page to copy one setup prompt into your own frontier agent. Choose the customer prompt to estimate Lithi Compute pricing or the provider prompt to estimate Mac earnings. Open the prompt dialog titled **Copy and paste this prompt into your agent.** Select **Copy prompt**. Review the prompt before you approve setup, quoting, execution, or a persistent change. Keep raw history, transcripts, secrets, and API keys in their approved local boundary. ## Copy the customer savings prompt Customer prompt version: `compute-public-assessment-2026-09-02-v1` Copy the complete block. It tells your frontier agent to review your authorised local work as a free assessment only, identify up to three bounded Compute-fit candidates with a current price preview each, and stop before any account, key, or paid-work action until you separately approve setup. ```text # Find where Lithi Compute fits You are the frontier coding agent I already use. Work locally and outcome-first. - Inspect only authorised local work, natural units, data boundaries, safety, and the current official Lithi Compute documentation. Keep raw history, transcripts, secrets, and keys here. - Identify up to three strongest bounded workload candidates that fit Lithi Compute, or explain why none fit. For each, show the outcome, natural unit, data fit, assumptions, current price preview, and uncertainty. Say **No like-for-like benchmark** when units differ. - Treat this as a free assessment only. Do not create an account, install anything, change persistent configuration, request or handle an API key, authenticate, reserve budget, call paid work, or contact a provider. - Stop after the fit and quote previews. If setup or execution would help, describe the next step and ask whether I separately want setup guidance; never continue into setup without that approval. - Do not call lithi.run or lithi.batch. No paid work is authorised by this prompt. Never print, copy, log, persist, or request secrets or keys. Return the findings clearly, with up to three candidates and a concise next action. ``` When setup is approved, use the secure Compute account flow to [Get my API key](/portal/compute/sign-up). Enter it yourself in the approved client or protected configuration. Never paste it into chat, a prompt, a log, a clipboard, or a repository. ## Copy the provider earnings prompt Provider prompt version: `agent-network-provider-earnings-2026-08-31-v2` Copy this separate block only when you provide approved Mac capacity. It keeps earnings as a forecast and stops before provider-visible action, deployment, or paid work without explicit approval. ```text # Provider fit and earnings handoff You are helping me estimate Mac earnings for Lithi inside the frontier agent I already use. ## Review - Study actual authorised workflows, work types, capacity, and fleet constraints locally. - Discover only management capabilities exposed to this agent and read current official provider security documentation. - Assess data profile, encryption suitability, unsafe cases, assumptions, and uncertainty. Treat earnings as forecasts, never guarantees. ## Boundaries - Never request raw credentials, secrets, tokens, or MDM credentials in chat; technical MDM access is not organisational authority. - Use current authorised quote and account data. No provider-visible action, deployment, or paid work before explicit approval. - Login never creates an account: guide explicit authenticated sign-in, create, or resume with official API/portal steps. - Solo needs no fake team approval. Organisation approval binds exact caller, quote, batch, policy version, expiry, and single-use identity, with a distinct approver where required. ## Output Return fit, bounded capacity assumptions, forecast range, strongest safe use cases, unsafe or unsupported cases, and the next explicit action. Stop when approval, auth, scope, safety, capacity, cost, or health changes. ``` ## Review provenance and calculate savings These fenced blocks project the versioned source in `src/lib/agent-network/setupPrompts.ts`. The customer block re-exports Compute's canonical `PUBLIC_ASSESSMENT_PROMPT` and is v1 dated 2026-09-02; the provider block remains v2. The source was retrieved on 2026-09-02 from the frozen source commit `e679d765119765411802edc6b7be534e961b99e5`, which converged the customer prompt onto that canonical source. Current authority IDs and references: - `docs/lithi_agent_docs/reference/architecture/COMPUTE_ESTATE_INDEX.json` - `scripts/ldcn/fixtures/compute-customer-capability-matrix.v1.json` - `docs/lithi_agent_docs/reference/LITHI_MASTER_GLOSSARY.yaml` - `contracts/compute/ComputeRevenueAllocationPolicyV1.yaml` - `src/app/(main)/docs/agent-resources/mcp/content.ts` - `contracts/api/openapi.yaml` - [OpenAI Codex MCP documentation](https://developers.openai.com/codex/mcp/) - [Claude Code MCP documentation](https://docs.anthropic.com/en/docs/claude-code/mcp) - [xAI MCP server documentation](https://docs.x.ai/build/features/mcp-servers) - [Google Antigravity MCP codelab](https://codelabs.developers.google.com/developer-knowledge-mcp-antigravity) - [OpenAI latest model guide](https://developers.openai.com/api/docs/guides/latest-model) Review the full prompt before you approve setup or any later action. When the review is complete, [Calculate My Savings](/compute)