Post by Johannes Tigges · 2026-08-28
====================================

Interessantes Paper zum Einfluss von AGENTS.md et al. Dateien. \
https://arxiv.org/pdf/2602.11988

> Wichtiger Quote aber auch:\
> Limitations and Future Work
> Programming languages The current evaluation is focused on Python. Since this
  is a language
> that is widely represented in the training data, detailed knowledge about
  tooling, dependencies, and
> other repository specifics might be present in the models' parametric
  knowledge, nullifying the effect
> of context files. Future work may investigate the effect of context files on
  more niche programming
> languages and toolchains that are less represented in the training data [8,
  27].
>
> Context files beyond task resolution
> In this work, we evaluate the impact of context files on task
> resolution rate. **However, other aspects of coding agent performance, such as
  code efficiency \[14]
> and security \[10], would be interesting directions for future work. Security,
  specifically, was found
> to be sensitive to prompt changes in prior work \[37].**

Hervorhebung eigene.

Im Prinzip also: Spitz, nützlich und was das Modell halt nicht selbst kann
  reinschreiben, nicht was Captain Obvious auch schon gewusst hätte.
  Whowouldathought. Klingt wie bei Menschen...😆

Tags: LLM, Aiassistedcoding, AI, research, contextengineering

Likes: 0 · Reposts: 2 · Bookmarks: 0 · Reactions from other networks: 2
  (@academia@newsmast.community shared this, @ai@channel.org shared this) ·
  Already counted in the numbers above.

--
vutuv agent document · type: post · schema_version: 3
generated_at: 2026-08-28T13:50:41Z
canonical:
  https://vutuv.de/johannes_tigges/posts/01a04837-bbc2-7aac-898e-afae8dbb2823