Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.thespawn.io/llms.txt

Use this file to discover all available pages before exploring further.

This pipeline adapts the Mandate docs playbook: separate Mintlify repo, explicit source map, parallel subagents, per-page writer packets, isolated QA, and content-quality gates.

​
Pipeline

  1. Inventory source truth from app routes, controllers, tests, the CLI repo, the public skill mirror, live endpoints, and deployment settings.
  2. Define developer journeys before information architecture, because a protocol-first sidebar hides the first successful action.
  3. Assign subagents by surface: CLI/MCP, API/ERC-8004/x402, skill/public docs, pipeline recovery, and DevEx scoring.
  4. Give each writer a packet with the customer job, sources, constraints, verified example, page contract, and cross-links.
  5. Build pages in the Mintlify repo and keep source-heavy notes in reference pages instead of spreading uncertainty through quickstarts.
  6. Run smoke checks against live examples, especially Social Intel API base:29382 and the hosted spawnr MCP runtime.
  7. Run Mintlify locally so broken navigation and MDX errors are caught before deployment.
  8. Run the DevEx scorecard and high-quality-content-writer gates, then treat failures as either content issues or code-block false positives.
  9. Iterate until the score is acceptable and every P0 journey has a command, expected signal, failure branch, and next step.
  10. Push the docs repo and wire the main product to the new docs domain.

​
Writer packet template

You are writing one The Spawn docs page.

CUSTOMER JOB:
<what the developer is trying to finish>

PRIMARY READER:
<agent consumer | agent builder | service provider | monetizer | quality improver>

SOURCES:
<exact files, commands, live endpoints, and prior notes>

VERIFIED EXAMPLE:
<copy-paste command and expected signal>

CONSTRAINTS:
- Start from the user's job, not protocol terminology.
- Keep x402, ERC-8004, MCP, wallet, and on-chain details progressive.
- Use exact endpoints and auth headers from code.
- Include failure recovery that names the likely broken surface and the next command to run.
- End with next steps that move to a neighboring job, not a generic conclusion.
- No hype, filler, or unsupported claims.

PAGE CONTRACT:
- Mintlify frontmatter with title and description.
- Include one runnable example where the product surface makes that possible.
- Include one expected response shape or checkpoint so the reader can tell whether the step worked.
- Link to adjacent pages that explain the next likely job.

​
Final implementation prompt

Take the SwiftAdviser/thespawn-docs Mintlify repo and build a developer-adoption docs system for The Spawn.

Goal: make a new agent developer successful on their first run, then guide them into deeper product paths: spawnr MCP, spawnr CLI, quickstarts, creating and managing agents with agent0/ag0 SDK or viem, ERC-8004 registration, x402 paid tools, The Spawn skill, quality scoring, wallet funding, and community contribution.

Start from customer jobs, not protocol names:
1. Find and hire a working agent.
2. Connect an AI client to The Spawn.
3. Publish an ERC-8004 agent.
4. Let agents use my service.
5. Charge for tool calls.
6. Improve my quality score.
7. Use The Spawn skill inside coding agents.

Use subagents aggressively:
- Source-truth scout: map routes, controllers, tests, CLI behavior, and live endpoint evidence.
- CLI/MCP scout: run first-run search/show/hire dry-run and direct MCP checks.
- API/ERC/x402 scout: verify endpoints, auth, quality scoring, payment challenge contract, and registration split.
- Skill scout: compare public skill files with docs and flag drift.
- Page writers: one journey or reference area each, using writer packets.
- DevRel reviewer: act as a DevRel lead with 10 years of developer adoption experience; score every journey with a written rubric.
- Content reviewer: run high-quality-content-writer on every changed page and require clean prose.

Quality bar:
- Every first-run journey must have a working command, expected signal, failure branch, and next step.
- Hide complexity without hiding value: explain x402, ERC-8004, on-chain agents, and MCP only after the reader has a concrete result.
- Use real examples, especially Social Intel API `base:29382`, only if the live checks pass.
- Document known limits and drift instead of smoothing over them.
- Wire docs.thespawn.io from the main thespawn.io product.
- Leave a QA report with commands, scores, blockers, and deployment state.

​
Review prompt

Review these docs as a DevRel lead with 10 years of experience launching developer platforms.

Score each journey from 0 to 100 using:
- First-run success: 30
- Source accuracy: 20
- Complexity hiding: 15
- Failure recovery: 15
- Mintlify structure: 10
- Community and feedback: 10

List P0/P1/P2 issues with file paths and exact fixes. Then identify the one change most likely to improve developer adoption this week.

​
Next