trae-agent

Select the best repository patch for a GitHub issue via candidate search and majority voting.

43|2|Updated Mar 26, 2026
One-click install
npx skills add https://github.com/curated-skills/LinguaClaw --skill trae-agent-curated-skills
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: trae-agent
Source: https://github.com/curated-skills/LinguaClaw/tree/main/harnesses/artifacts/trae-agent
Command: npx skills add https://github.com/curated-skills/LinguaClaw --skill trae-agent-curated-skills

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

It solves unreliable “single-edit” bug fixing by treating a GitHub issue as a repository search over many independently sampled patch candidates and then selecting the best one via explicit pruning and voting.

Core Features & Use Cases

  • Independent candidate generation (fresh starts): generates C1–C10 candidates without reusing intermediate reasoning, command history, or partial patches across candidates.
  • Explicit pruning, selection, and majority-vote discipline: prunes weak/duplicate candidates, ranks survivors, and performs grouped selection with majority voting (default budget) instead of one-shot picking.
  • Clean final revalidation: reruns reproduction steps and targeted tests in a final validation pass, recording why the chosen winner survived.

Quick Start

Ask an agent to run the trae-agent harness on the target repository issue, using the default budget of 10 independently generated candidates with grouped majority-voting selection and a clean final revalidation of the chosen winner.

Frequently Asked Questions about trae-agent

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

FAQPage Schema
How do I fix GitHub issues using multi-candidate patch selection?

To fix GitHub issues, you run a TRAE-style candidate search that generates up to 10 independent full patch attempts, prunes weak duplicates, and uses majority voting to select the best repository fix. This ensures reliable bug repair instead of relying on a single iterative edit.

Why does single-edit bug fixing fail on repository issues?

Single-edit bug fixing fails because iterative in-session edits are not a faithful proxy for diverse candidates. Without fresh-start isolation per patch attempt, the reasoning history biases the output and prevents finding the optimal repository patch.

How does majority voting work for software bug repair workflows?

Majority voting for software bug repair works by grouping independently sampled patch candidates, pruning weak or duplicate attempts, ranking survivors, and selecting a grouped winner. A final validation pass then reruns targeted tests to confirm the chosen patch survives.

Does test revalidation happen after selecting the winning patch?

Test revalidation does happen after selection by running a clean final validation pass. It reruns reproduction steps and targeted tests on the chosen winner, recording structured evidence in a ledger to explain why the selected patch survived.

Can I adjust the candidate search budget for repository fixes?

You can adjust the candidate search budget for repository fixes, which defaults to 10 independently generated candidates. This budget controls the number of fresh-start patch attempts before the pruning and grouped majority-voting selection stages begin.