ryfineZ/codex-session-patcher

A lightweight Python tool to clean AI refusal responses from Codex CLI session files

What it solves

AI coding tools like Codex, Claude Code, and OpenCode often refuse requests involving security-sensitive operations, which interrupts the conversation flow during security testing, CTF (Capture The Flag) competitions, or penetration testing. This project prevents these interruptions by cleaning up refusal messages from session histories and injecting specialized contexts to reduce the likelihood of refusals from the start.

How it works

The tool employs two primary strategies:

  1. Session Cleaning: It scans session logs (JSONL or SQLite formats) for refusal patterns using a two-tier detection strategy. It can then replace these refusals with cooperative content, either using a safe default text or by calling an LLM (via OpenAI, Ollama, or OpenRouter) to generate context-aware replacements. It also erases reasoning/thinking blocks to clean the output.
  2. Prompt Injection: It modifies the configuration of the AI tools to inject security-testing contexts. For Codex, it creates specific profiles or modifies global settings; for Claude Code and OpenCode, it creates dedicated CTF workspaces with specialized instruction files (like CLAUDE.md or AGENTS.md).

Who it’s for

Security researchers, CTF players, and penetration testers who use AI coding assistants to help with vulnerability analysis, exploit development, or reverse engineering.

Highlights

  • Multi-Platform Support: Works with Codex CLI, Claude Code, and OpenCode.
  • AI-Powered Rewriting: Uses LLMs to rewrite both session refusals and user prompts to make them more acceptable to the AI.
  • Visual Management: Includes a Web UI for session list management, side-by-side diff views of cleaned sessions, and prompt editing.
  • Automated CTF Workflows: Provides built-in templates for various security tasks (Web, Pwn, Reverse, Crypto, etc.) that automatically match the task type.
  • Safety Features: Automatically creates backups before cleaning and supports one-click restoration.

Related

  • Dispatch
    Maximizing Claude Code Session Value and Token Efficiency

    Optimizing Claude Code sessions requires strategic use of prompt caching, context management, and subagents to reduce token costs and prevent cache misses.

  • Project
    dob323/session-kit

    Session Kit is a local command‑line dashboard that lists all your Claude Code and Codex AI coding sessions, shows their state (e.g., waiting for input), and lets you jump to any session by typing its stable number. Sessions survive terminal or SSH disconnects, can use different provider accounts, and are protected by a safety model that re‑verifies the exact provider process before acting. It installs via a verified release, runs on Linux (systemd) or macOS 14+, and stores all data locally with no telemetry.

  • Project
    severity1/claude-code-prompt-improver

    A prompt optimization plugin for Claude Code that injects context and clarifies vague prompts to ensure better first-attempt results.

  • Dispatch
    OpenCode 0.0 – Why the Popular AI Coding Agent Is a Security and Usability Nightmare

    OpenCode’s design choices cause severe prompt‑cache misses, broken permission handling, and remote‑first defaults that make it both annoying to use and alarmingly insecure.

  • Project
    ericlitman/open-pstack

    A plugin for Claude Code and Codex that implements a rigorous engineering framework for coding agents, emphasizing verification, multi-model review, and structured workflows.