gtg 0.7.5

Deterministic PR readiness detection for AI coding agents

Good To Go

Deterministic PR readiness detection for AI coding agents

"Is the PR ready to merge?" — Finally, a definitive answer.

Documentation · PyPI · Contributing

---

Why Good To Go?

AI agents can write code, fix bugs, and respond to reviews. But they all struggle with one question: "Am I done yet?"

• CI is running... check again... still running...
• CodeRabbit left 12 comments — which ones are blocking?
• Reviewer wrote "consider X" — is that a request or a suggestion?
• Threads are unresolved — but the fix is already pushed

Good To Go answers this definitively:

gtg 123

OK PR #123: READY
   CI: success (5/5 passed)
   Threads: 3/3 resolved

One command. One status. No guessing.

How It Works

flowchart LR
    PR[Pull Request] --> GTG[gtg]
    GTG --> CI[CI Status]
    GTG --> Comments[Comment Analysis]
    GTG --> Threads[Thread Resolution]

    CI --> Status{Status}
    Comments --> Status
    Threads --> Status

    Status --> READY[✓ READY]
    Status --> ACTION[! ACTION_REQUIRED]
    Status --> UNRESOLVED[? UNRESOLVED_THREADS]
    Status --> FAILING[✗ CI_FAILING]

Good To Go combines three analyses:

Analysis	What It Does
CI Status	Aggregates all checks into pass/fail/pending
Comment Classification	Identifies actionable vs. informational comments
Thread Resolution	Tracks which discussions are truly blocking

Intelligent Comment Classification

Not all comments need action. Good To Go classifies each one:

Classification	Examples	Action
ACTIONABLE	"Critical: SQL injection vulnerability"	Must fix
NON_ACTIONABLE	"LGTM!", nitpicks, resolved items	Ignore
AMBIGUOUS	"Consider using X", questions	Human review

Built-in support for: CodeRabbit, Greptile, Claude, Cursor/Bugbot

Quick Start

# Install
pip install gtg

# Set GitHub token
export GITHUB_TOKEN=ghp_...

# Check a PR (auto-detects repo)
gtg 123

# Explicit repo
gtg 123 --repo owner/repo

Output Formats

Text (Human-Readable)

gtg 123 --format text

!! PR #456: ACTION_REQUIRED
   CI: success (5/5 passed)
   Threads: 8/8 resolved

Action required:
   - Fix CRITICAL comment from coderabbit in src/db.py:42

JSON (For Agents)

gtg 123 --format json

{
  "status": "ACTION_REQUIRED",
  "action_items": ["Fix CRITICAL comment from coderabbit in src/db.py:42"],
  "actionable_comments": [...],
  "ci_status": {"state": "success", "passed": 5, "total_checks": 5}
}

Exit Codes

Default (AI-friendly) — parse the JSON for details:

Code	Meaning
0	Any analyzable state
4	Error

With -q or --semantic-codes — for shell scripts:

Code	Status
0	READY
1	ACTION_REQUIRED
2	UNRESOLVED
3	CI_FAILING
4	ERROR

Use as CI Gate

Make gtg a required check to block merges until PRs are truly ready:

# .github/workflows/pr-check.yml
- name: Check PR readiness
  run: gtg ${{ github.event.pull_request.number }} --semantic-codes

See USAGE.md for full workflow setup.

For AI Agents

import subprocess
import json

result = subprocess.run(
    ["gtg", "123", "--format", "json"],
    capture_output=True, text=True
)
data = json.loads(result.stdout)

if data["status"] == "READY":
    print("Merge it!")
else:
    for item in data["action_items"]:
        print(f"TODO: {item}")

Or use the Python API:

from goodtogo import PRAnalyzer, Container

container = Container.create_default(github_token="ghp_...")
analyzer = PRAnalyzer(container)
result = analyzer.analyze("owner", "repo", 123)

State Persistence

Track handled comments across sessions:

gtg 123 --state-path .goodtogo/state.db  # Remember dismissed comments
gtg 123 --refresh                         # Force fresh analysis

Documentation

• Landing Page — Philosophy and vision
• USAGE.md — Complete CLI reference
• CONTRIBUTING.md — Development guide

License

MIT License — see LICENSE

---

Made with Claude Code
by David Sifry