Motivation

trajgen turns verified SWE tasks into agent trajectories and ready-to-train SFT data.

It is the trajectory-generation stage of the SWE-Lego-Live pipeline, sitting between task curation (swegen) and model training (sft, rl). It rolls a coding agent out across verified tasks in reproducible containers, captures every rollout, and converts the good ones into SFT data — all driven by a single config around a pinned Harbor checkout.

swegen ─▶ trajgen ─▶ sft ─▶ rl

trajgen provides:

• Manifest-gated consumption — only swegen-verified tasks run, and a ledger guarantees each is rolled out exactly once
• A per-job LiteLLM proxy fronting your model API with OpenAI- and Anthropic-compatible endpoints
• Containerized rollouts at scale via Harbor, with configurable concurrency, retries, and timeouts
• Pluggable agent scaffolds — Claude Code, OpenCode, OpenHands SDK, and Terminus-2 pre-integrated
• Automatic SFT conversion to LLaMA-Factory format, with rule- and LLM-based quality scoring
• A live progress dashboard you can preview locally or publish to Cloudflare Pages

Where to go next

• Getting Started — install the dependencies and run your first job
• Core Concepts — tasks, trajectories, jobs, scaffolds, and SFT data
• Run Jobs — prepare tasks, start the proxy, and roll the agent out
• SFT Data — convert trajectories into training datasets
• Dashboard — monitor progress locally or online