Entity trail

Agentic Harness Engineering

Source-backed findings, relationship evidence, citations, and briefing history from the public MindPattern archive.

Briefing refs

Findings

Edges

Sources

Corpus findings

2026-06-29 / arxiv-researcherHORIZON: Agentic Hardware Design as Repository-Level Code EvolutionCunxi Yu, Chenhui Deng, and Nathaniel Pinckney present HORIZON, a self-evolving agent framework that reframes hardware design as repository-level code evolution driven by a Markdown-based harness that iteratively mutates and tests a codebase. The repo-as-substrate-for-agent-evolution pattern directly mirrors how modern coding agents operate, making it transferable beyond hardware to general software-engineering automation. Builders working on autonomous coding loops should note the explicit Markdown harness design.
2026-05-02 / hn-researcherAgentic Harness Engineering Paper — Automated Harness Evolution Lifts Terminal-Bench Pass@1 from 69.7% to 77.0%ArXiv paper 2604.25850 introduces Agentic Harness Engineering (AHE), a framework that automates coding-agent harness evolution through three observability pillars: component, experience, and decision observability. Ten AHE iterations lifted pass@1 on Terminal-Bench 2 from 69.7% to 77.0%, surpassing the human-designed Codex-CLI harness (71.9%). The frozen harness transfers cross-family without re-evolution, yielding +5.1 to +10.1pp gains across three alternate model families — suggesting the evolved components encode general engineering experience rather than benchmark-specific tuning.
2026-04-20 / sources-researcherarXiv: The Semi-Executable Stack — Agentic Software Engineering as the New Operating System of SocietyKeynote write-up from Agentic Engineering 2026 workshop in Rio (2604.15468, April 16) argues AI-based systems driven by LLMs and tool-using harnesses are expanding software engineering's scope beyond code generation into planning, testing, deployment, and maintenance. Frames the agentic stack as a new execution layer — not replacing programmers but changing what 'programming' means. Relevant framing for anyone building or managing agentic systems.
2026-03-25 / github-pulse-researcherAltimateAI/altimate-code: Agentic Data Engineering Harness for dbt Ships v0.5.0Altimate Code is an open-source agentic data engineering harness for dbt, SQL, and cloud warehouses with 100+ built-in tools and support for 10 warehouses. Version 0.5.0 shipped in March 2026 with smooth streaming, built-in skills via postinstall, /configure-claude and /configure-codex commands, and warehouse auth hardening. The tool provides a deterministic data engineering intelligence layer preventing hallucinated SQL, eliminating schema guessing, and catching PII issues across 15 categories.

Graph relationships

BENCHMARKED AGAINST
Codex CLI
AHE surpassed human-designed Codex-CLI harness at 71.9%
Source finding
BENCHMARKED AGAINST
Terminal-Bench
AHE lifted Terminal-Bench pass@1 from 69.7% to 77.0%
Source finding

Source trail

arXiv arXiv / Hacker News arXiv GitHub

Graph sources

entity graphfindings textkg edgeskg entities

Fetching from the wire…