contribute-to-eliza

Implements or independently reviews scoped elizaOS GitHub issues and pull requests with verified evidence.

1|Updated Aug 5, 2026
One-click install
npx skills add https://github.com/Solizardking/Solana-Robotics-Kit --skill contribute-to-eliza-solizardking
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: contribute-to-eliza
Source: https://github.com/Solizardking/Solana-Robotics-Kit/tree/main/eliza/packages/skills/skills/contribute-to-eliza
Command: npx skills add https://github.com/Solizardking/Solana-Robotics-Kit --skill contribute-to-eliza-solizardking

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) components.

What problem does it solve? Contributing to a large open-source repository like elizaOS requires disciplined workflows: selecting unclaimed work, implementing changes with real tests, capturing verifiable evidence, and safely reviewing untrusted pull request code. This Skill enforces that entire contribution lifecycle with strict security isolation and attribution rules. ## Core Features & Use Cases - Issue Completion (Mode A): Claim a scoped, contributor-ready issue, implement the change on a correctly prefixed branch, add real tests, and open a PR against develop with a complete evidence matrix. - Independent PR Review (Mode B): Inspect an open PR from a trusted control checkout, execute untrusted code only inside a disposable sandbox, and leave actionable review findings or authorized repairs. - Candidate Inventory & Compliance Audits: Run scripts/live-report.mjs to paginate GitHub issues and PRs, filter claimed, sensitive, or bot-authored work, and audit model-disclosure and evidence-row gaps. - Use Case: An autonomous agent selects an unclaimed good first issue in elizaOS/eliza, implements the fix, runs bun run verify, captures logs and screenshots per the evidence rubric, and opens a PR with a signed provider/model attribution footer. ## Quick Start Ask the agent to use the contribute-to-eliza skill to find one unclaimed contributor-ready issue in elizaOS/eliza and take it through implementation, testing, and PR submission.

Frequently Asked Questions about contribute-to-eliza

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I contribute to elizaOS issues with an AI agent?

Run the live-report script to list unclaimed contributor-ready issues, claim one publicly with a CLAIMING comment and provider/model disclosure, implement the change on a feat or fix branch, add real tests, and open a PR against develop with the full evidence matrix.

How do I safely review an untrusted pull request on GitHub?

Fetch the exact PR head SHA without checking it out, inspect the diff from a trusted control checkout, then execute builds and tests only inside a fresh disposable container with no credentials, no network, and bun install --frozen-lockfile --ignore-scripts.

What evidence is required for an elizaOS pull request?

The PR template requires seven rows: before and after screenshots, a walkthrough video, backend logs, frontend logs, a real-LLM trajectory, and domain artifacts. Each row needs concrete inline evidence or a specific N/A reason; a checked box alone does not qualify.

Can I run PR tests that need network access or credentials?

Only with explicit operator approval in a separate single-use sandbox with allowlisted egress and an ephemeral least-privilege credential created for that test. Never pass through the agent's normal gh token, and revoke the test credential immediately afterward.

Why does the contribution workflow stop before posting to GitHub?

Posting stops when the exact AI provider or model identifier is unavailable, or when skill revision provenance is missing, dirty, or mismatched. These are hard stop conditions; the agent must ask the operator rather than substitute placeholders or guessed values.