Ghost Hunter
A bot that automatically reproduces GitHub bug reports: it reads the issue with an AI, tries the steps in a disposable Docker sandbox, and posts back exactly what happened.
🔗 Visit Ghost Hunter
Description
"Can't reproduce" is one of the most frustrating replies a bug report can get — someone has to manually follow the reporter's steps, often on a machine that doesn't match theirs. Ghost Hunter automates that chore: comment "reproduce" on a GitHub issue, and a bot reads the report with an AI, tries the steps in a disposable sandbox, and posts back exactly what happened.
Ghost Hunter is a Python/FastAPI webhook service triggered by a "bot/reproduce" comment on a GitHub issue. It uses an LLM via OpenRouter to parse the issue into concrete reproduction steps, runs them inside an isolated Docker container, and posts the resulting crash log back to the issue, with HMAC SHA-256 webhook verification, SQLite-based dedup to avoid repeat runs, and retry logic (up to 3 attempts) that feeds failed attempts back to the LLM as context. It supports GitHub auth via Personal Access Token or GitHub App, and Smee.io for local webhook testing. Requires Docker Desktop running. MIT licensed.
💬 Our review
The short version: Ghost Hunter automates a specifically tedious maintainer chore, bug reproduction, by handing the LLM the reading comprehension and Docker the actual sandboxed execution.
Unlike general coding agents that try to fix the bug directly, Ghost Hunter's scope is narrower and more reliable by design: it only reproduces, using an isolated container so a bad repro script can't do real damage. The weak link is the LLM parsing step: the project's own docs flag that free-tier OpenRouter models can return malformed JSON, so reproduction quality depends on which model you point it at. Free and open-source, worth trying on a repo that gets a lot of vague bug reports; skip it if your issues rarely include enough detail to reproduce in the first place, since garbage-in still means garbage-out here.
💰 Pricing
📊 Global score
🤖 AI-enriched data
Logiciel gratuit (MIT), nécessite une clé API OpenRouter pour les appels LLM
Pros
Automatise une tâche répétitive et frustrante (reproduction de bugs)
Exécution isolée dans Docker, aucun risque pour la machine hôte
Vérification cryptographique des webhooks, plusieurs méthodes d'authentification GitHub
Cons
Qualité dépendante du modèle LLM choisi, les modèles gratuits peuvent produire du JSON mal formé
Nécessite Docker Desktop actif en permanence
Le proxy Smee.io peut être instable en cas de coupure réseau