Reproduction package for "Harness as a Language: A Minimalist Agent Framework With Maximal Expressivity" — ten agent arms across AppWorld and StuLife
-
Updated
Oct 10, 2026 - Python
#
appworld
Here are 11 public repositories matching this topic...
DAEDALUS: Bootstrapping Agent Memory from Self-Generated Tasks. Code, paper and blog post.
-
Updated
Oct 19, 2026 - Python
Agentic Soft-Token Compression (ASTC): training-free in-model context compression for long-horizon LLM agents.
-
Updated
Oct 15, 2026 - Python
[arXiv] ExpHarness: Model-Agnostic Experience Learning through a Trainable Harness
reinforcement-learning retrieval harness alfworld llm-agents agent-memory experience-learning appworld frozen-llm experience-graph
-
Updated
Oct 4, 2026 - Python
Minimal AppWorld agent for auto-agent optimization experiments — target agent that gets autonomously improved
-
Updated
Mar 25, 2026 - Python
Pi-Bench research workspace for proactive-agent evaluation, memory-backend integration, experiment orchestration, and diagnostics.
docker personal-assistants research-infrastructure llm-evaluation agent-memory agent-benchmark proactive-agents appworld
-
Updated
Sep 1, 2026 - Python
Harness-localized credit assignment and certificate-guided RL for AppWorld agents
-
Updated
Jul 11, 2026 - Python
Controlled comparisons of ordinary tools and Code Mode on agent recovery and conditional batching, with saved traces, offline verification, and live runners.
-
Updated
Oct 25, 2026 - JavaScript
Experimental code and selected traces for automatic task discovery and verification
-
Updated
Aug 31, 2026 - Python
Task-conditioned tail reliability for tool-using agents under equivalent interfaces
-
Updated
Oct 24, 2026 - Python
A2A green-agent orchestrator for evaluating agents on the AppWorld benchmark, built on the AgentBeats SDK
-
Updated
Jul 1, 2026 - Python
Add this topic to your repo
To associate your repository with the appworld topic, visit your repo's landing page and select "manage topics."