OpenForgeRL enables end-to-end training for real-world agent harnesses
The framework separates training from inference, logs harness calls, and scales rollouts in containers — posting 31.7/55.9 pass@3 on ClawEval and 37.7 on OSWorld-Verified. A related study reports 4x faster multi-turn distillation, while llama.cpp and jCodeMunch keep local and low‑token workflows practical.
The business desk's take → Samsung signs $200B Broadcom AI chip pact to power its foundry push30-second briefing
- OpenForgeRL enables end-to-end RL on real harnesses.
- ReOPD reuses prefixes, 4x faster per rollout.
- llama.cpp and jCodeMunch ship stability fixes.
One-Line Summary
Agent training steps into real-world stacks: OpenForgeRL enables end-to-end training on actual harnesses, ReOPD speeds multi-turn distillation 4x, and repos refine local and low-token workflows.
Research Papers
OpenForgeRL trains harness-native agents across environments
OpenForgeRL is a framework that lets teams train AI agents inside the same orchestration “harnesses” they actually use—like tool-using coders or GUI browser agents—without rewriting infrastructure. It inserts a lightweight proxy that serves the harness’s model calls while recording them for a standard Reinforcement Learning (RL) pipeline, and a Kubernetes orchestrator that runs each rollout in its own remote container. Using only hundreds to a few thousand tasks, the authors report OpenForgeClaw at 31.7 pass^3 and 55.9 pass@3 on ClawEval and 33.7 on QwenClawBench; OpenForgeGUI reaches 37.7 on OSWorld-Verified, 63.0 on Online-Mind2Web, and 72.3 on WebVoyager. 1
The practical move here is decoupling training from inference: the proxy keeps production-like harness behavior intact while turning those calls into training data for a conventional RL codebase. The containerized rollouts mean researchers can scale experiments across “any harness in any environment” without hacking their training loop to mimic complex, stateful, multi-process inference. 1
Beyond scores, the paper analyzes how harness choice and RL shaping change behavior. Some harnesses are substantially harder to learn; RL improves reliability markers like self-verification, tool coverage, and completing multi-step plans, while error recovery remains a weak spot to watch. The sample efficiency suggests a credible path to training agents directly where they will run. 1
Context: modern agents rely on elaborate inference harnesses (e.g., Claude Code, Codex, OpenClaw) for multi-turn reasoning and tool use, but typical open Supervised Fine-Tuning (SFT) and RL stacks cannot natively express that complexity. OpenForgeRL offers a general adapter to bridge research training with real deployments. 1
ReOPD cuts multi-turn distillation cost with replayed prefixes
Replayed-Prefix On-Policy Distillation (ReOPD) trains a smaller student to imitate a teacher across multi-turn agent interactions without repeatedly running the environment. For agentic tasks with a Large Language Model (LLM), it reuses pre-collected teacher trajectories as “replayed prefixes,” lets the student act at selected steps, and has the teacher provide dense per-step supervision offline—an alternative to costly fully online On-Policy Distillation (OPD). 2
The authors surface a “prefix trap”: making histories more student-on-policy improves relevance but risks querying the teacher where its target is unreliable, creating a two-sided distribution shift between student occupancy and teacher reliability. ReOPD treats this as a reliability-aware prefix distribution problem and implements a simple step‑decaying sampling schedule emphasizing early, lower-shift prefixes. 2
Across math-with-Python and search environments at multiple teacher/student scales, ReOPD preserves or improves OPD-level accuracy, uses zero tool calls during student training, and is at least 4x faster per rollout than standard OPD—turning expensive agent-environment interaction into a reusable offline resource. 2
Open Source & Repos
llama.cpp updates keep local C/C++ inference accessible
llama.cpp is a C/C++ project for running Large Language Model (LLM) inference locally with minimal dependencies. The latest b10107 release (Jul 24) notes a Windows crash fix for the Hexagon backend when op_poll is enabled, alongside active release artifacts and CI workflows for quick setup across platforms. 3
With MIT licensing, prebuilt binaries, server and Docker workflows, and a simple command-line, llama.cpp remains a straightforward path to try on-device text generation or privacy-first prototypes without heavy frameworks. It’s a practical base for teams testing local inference before committing to larger stacks. 3
jCodeMunch MCP slashes code browsing tokens via symbol search
jCodeMunch MCP is a Model Context Protocol (MCP) server that pulls precise GitHub code snippets at the symbol level using tree-sitter Abstract Syntax Tree (AST) parsing—so agents read only the functions they need instead of entire files. The project claims 95%+ token reduction for code exploration and provides live telemetry. 4
The repo reports 335B+ tokens saved, 48,000+ developers, and $1.69M+ in avoided AI spend, with a benchmark citing 99.6% average token reduction (15 tasks, 3 repos). The v1.108.184 release (Jul 26) closes a defect around a ‘negative_evidence’ path and missing verdict metadata—incremental hardening for production agent workflows in tools like Claude Code and Cursor. 4
Why It Matters
Today’s papers tackle the hardest part of agents: making training and distillation fit the messy, stateful environments where tools, browsers, and GUIs actually run. OpenForgeRL points to end-to-end training directly on real harnesses, while ReOPD shows you can reuse past interactions to train faster and cheaper without live tool calls. 1
On the tooling side, llama.cpp keeps local inference approachable, and jCodeMunch uses symbol-level retrieval to curb token use. Together, they reduce friction and cost across the agent stack—from data collection and training to deployment and day‑to‑day coding support. 3
This Week to Try
- Run a local LLM with llama.cpp: grab the latest release binaries and follow the README to generate text on your machine. https://github.com/ggml-org/llama.cpp
- Cut IDE token spend with jCodeMunch MCP: install the server per the repo guide and enable symbol‑level retrieval in your MCP‑compatible editor. https://github.com/jgravelle/jcodemunch-mcp
Terms in This Article
- Token
A token is the basic unit an LLM uses to represent and process text instead of reading raw characters or full words directly.
- RL
Reinforcement Learning is a machine learning paradigm in which an agent improves its decision policy by interacting with an environment to maximize cumulative reward, formalized as a sequential decision-making problem—typically a Markov Decision Process—where feedback is given as scalar rewards following actions.
- Inference
Inference is the execution phase where a trained model receives new inputs and computes predictions, classifications, or generated outputs. For LLMs, prompts are tokenized, processed through prefill and decode, and emitted as tokens while latency, throughput, and memory usage become core operating metrics.
Two desks on this story
OpenForgeRL enables end-to-end training for real-world agent harnesses
Samsung signs $200B Broadcom AI chip pact to power its foundry push
The agreement spans memory, foundry and advanced packaging, including sub-2nm production for Broadcom communications chips. It signals rising demand for custom accelerators and gives Samsung leverage as it courts top-tier customers.
Comments (0)