Skip to content

LP-0019.md

Latest commit

 

History

History
150 lines (111 loc) · 7.07 KB

LP-0019.md

File metadata and controls

150 lines (111 loc) · 7.07 KB
Raw

LP-0019: AI-Powered Logos Ecosystem Onboarding Toolkit [Open]

Status: Open Logos Circle: N/A

Overview

This prize funds a suite of open-source developer tools that lower the barrier to understanding and operating within the Logos ecosystem:

  1. A conversational AI assistant embedded in the Logos validator staking interface, answering questions about the Logos stack, testnet operations, and ecosystem participation — live, without a centralised AI backend requirement.

  2. An AI agent skill that installs, upgrades, and monitors a Logos testnet node from within any major AI coding assistant (Claude Code, GitHub Copilot, Gemini CLI, and others) with a single command, fetching live release metadata at runtime.

  3. A block indexer and REST API that ingests all canonical Logos testnet blocks, transactions, and proof-of-leadership key diagnostics into PostgreSQL, and exposes a structured HTTP API for explorer frontends.

Together these three components form an entry-point layer for the Logos ecosystem: a developer can ask what Logos is, get guided through running a node, and then observe their node's activity on-chain — all using open, self-hosted tooling.

Motivation

Logos is building privacy-preserving infrastructure. Growing the validator set and developer community requires tooling that meets people where they already are: in their AI assistants, on staking dashboards, and in their browsers. Today these entry points are either missing or require deep protocol knowledge to navigate.

This prize targets the onboarding gap directly: no new protocol modules are required, but the result meaningfully expands who can participate in the testnet and how quickly they can go from first contact to running a node and querying chain state.

A competitive prize is the right mechanism because the assistant training corpus, skill interface design, indexer schema, and API surface admit many valid approaches, and the best implementations will be discovered through competition.

Success Criteria

AI Documentation Assistant

  • A conversational assistant is live and accessible at a public URL, answering questions about the Logos stack, testnet participation, and node operations.
  • The assistant is embedded in or linked from a Logos-related validator or staking interface, making it discoverable to validator operators.
  • The assistant uses Logos documentation as its primary knowledge source and does not return factually incorrect information about the Logos protocol in standard queries.
  • The assistant is self-hostable: the repository includes deployment instructions for running the assistant independently, with no mandatory dependency on a specific hosted AI provider.

AI Agent Skill for Node Operations

  • The skill implements three commands: install (fresh node setup), update (breaking upgrade with state wipe), and status (sync mode, peers, block height, wallet balance).
  • The skill fetches the current stable Logos release from the GitHub API at runtime; it correctly filters out pre-releases, release candidates, and drafts. No hardcoded version strings.
  • The skill is installable via npx with a single command. No prerequisites beyond Node.js.
  • The skill works with at least 3 distinct AI agent platforms from: Claude Code, GitHub Copilot, Gemini CLI, OpenAI Codex CLI, opencode, OpenClaw, or equivalent.
  • A documented SKILL.md defines the skill interface. A third party can implement a compatible skill without reading the implementation.
  • If the GitHub API is unavailable, the skill falls back to the official Logos docs and surfaces a clear message to the operator.

Block Indexer and Explorer API

  • The indexer ingests all canonical Logos testnet blocks, transactions, and proof-of-leadership key diagnostics from a running Logos node into PostgreSQL.
  • The indexer correctly derives canonical block heights from the cryptarchia tip chain (Logos v0.1.2 does not expose stable per-block height in the REST API).
  • A REST API exposes at minimum: blocks (paginated, filterable by finality and canonical status), transactions (paginated), leader-key diagnostics, and a stats summary endpoint.
  • The indexer and API are deployable via Docker Compose with a single command.
  • A public deployment of the indexer API is live and queryable during evaluation.

General

  • All three components are in public repositories licensed under MIT or Apache-2.0.
  • Each component includes a README with setup, configuration, and usage instructions.

Scope

In Scope

  • Conversational AI assistant for Logos documentation and ecosystem onboarding.
  • AI agent skill for Logos node installation, upgrade, and status monitoring.
  • Block indexer and REST API for Logos testnet chain data.
  • Self-hosting and deployment documentation for all three components.

Out of Scope

  • Logos Core protocol modules, LEZ programs, or ZK proof generation.
  • Wallet UI or token management interfaces.
  • Windows or macOS node installation support (Logos ships Linux x86_64 binaries only).
  • Full-text search or semantic indexing of transaction content.

Prize Structure

  • Total Prize: TBD (determined by Logos team)
  • Effort: Medium

Eligibility

Open to any individual or team. Submissions must be original work. Teams must hold the rights to all submitted code and agree to license it under MIT or Apache-2.0.

Submission Requirements

  • Public repositories (MIT or Apache-2.0) for all three components.
  • Live, accessible URLs for the AI assistant and the block indexer API.
  • README files covering setup, deployment, and usage for each component.
  • GitHub issues filed for any Logos tooling problems encountered during development.

Evaluation Process

Submissions are evaluated first-come-first-served against the success criteria. Evaluators may verify live deployments directly via the submitted URLs and clone repositories to run locally.

The following policies apply to all prizes (see evaluation policies):

  • Submissions: each builder (or team) is allowed a maximum of 3 submissions per prize, with at most one submission/review per week.
  • Feedback: initial evaluation feedback is limited to a pass/fail indication against the success criteria.

Resources

Potential for Subsequent λPrizes

A follow-up prize may replace the assistant's static document corpus with a RAG-backed MCP server (logos-rag-mcp) that indexes Logos documentation automatically and serves context-aware answers to any compatible AI agent skill — eliminating manual updates when Logos documentation changes.