Skip to content

Wheels 2.x > 4.x Performance Comparison #2061

Wheels 2.x > 4.x Performance Comparison

Wheels 2.x > 4.x Performance Comparison #2061

Workflow file for this run

name: Wheels Bot — Cross-Framework Research
on:
issue_comment:
types: [created]
workflow_dispatch:
inputs:
issue-number:
description: 'Issue number to research'
required: true
type: string
permissions:
contents: read
concurrency:
group: wheels-bot-research-${{ github.event.issue.number || inputs.issue-number }}
cancel-in-progress: false
jobs:
research:
name: Cross-framework research
runs-on: ubuntu-latest
timeout-minutes: 30
# Phase 4 (auto-fire): runs from wheels-bot[bot] comments containing
# `wheels-bot:triage-class:framework-design`. workflow_dispatch is
# preserved for manual reruns.
# The bot-identity check is load-bearing: prevents humans from
# triggering research by quoting a marker in a comment reply.
if: |
vars.WHEELS_BOT_ENABLED == 'true'
&& (
github.event_name == 'workflow_dispatch'
|| (github.event_name == 'issue_comment'
&& github.event.comment.user.login == 'wheels-bot[bot]'
&& contains(github.event.comment.body, 'wheels-bot:triage-class:framework-design'))
)
env:
ISSUE_NUMBER: ${{ github.event.issue.number || inputs.issue-number }}
steps:
- name: Checkout develop
uses: actions/checkout@v6
with:
ref: develop
fetch-depth: 1
- name: Generate App token
id: app-token
uses: actions/create-github-app-token@v2
with:
app-id: ${{ secrets.WHEELS_BOT_APP_ID }}
private-key: ${{ secrets.WHEELS_BOT_PRIVATE_KEY }}
- name: Skip check
id: gate
uses: ./.github/actions/wheels-bot-skip-check
with:
target-type: issue
target-number: ${{ env.ISSUE_NUMBER }}
marker-pattern: 'wheels-bot:research:${{ env.ISSUE_NUMBER }}'
github-token: ${{ steps.app-token.outputs.token }}
- name: Run Research
if: steps.gate.outputs.skip == 'false'
uses: anthropics/claude-code-action@v1
with:
# When un-gated at Phase 4b, this workflow fires from wheels-bot[bot]
# comments containing the framework-design class marker. Allow that
# bot identity (and github-actions[bot] for consistency).
allowed_bots: 'wheels-bot[bot],github-actions[bot]'
anthropic_api_key: ${{ secrets.ANTHROPIC_API_KEY }}
github_token: ${{ steps.app-token.outputs.token }}
prompt: |
/research-frameworks ${{ env.ISSUE_NUMBER }}
# Model policy: judging gate = fable, coding stages = opus, janitorial = sonnet.
claude_args: |
--model claude-opus-4-8
--max-turns 400
--allowedTools "Bash(gh:*),Bash(git log:*),Bash(git show:*),Bash(git grep:*),WebFetch,WebSearch,Read,Grep,Glob"