Adds guardrails around delegating tasks to OpenAI Codex, including quota preflight and handoffs, git worktree isolation, per-turn checkpoints and restore, read-only cross-reviews, and honest failure reporting.
Enables Talon agents to delegate autonomous coding tasks to OpenAI Codex, supporting multi-step code operations with configurable sandboxing and approval policies.
Enables Claude Code to delegate bounded, read-only analysis and isolated patch proposals to the locally installed Codex CLI over MCP, with sanitized live status, revision-aware polling, and reviewable results.