by fishmingyu · Agent Tool · ★ 36
OrcaLoca 🔥 News! 🚀 Apr 5 2025: We support top-K retrieval mode inspired from LocAgent and CoSIL. Feb 18 2025: We update the support for Gemini series model through vertex API. OrcaLoca (previous named Orcar), an LLM agent framework that improves accuracy for software issue localization by integrating priority-based scheduling for LLM-guided action, action decomposition with relevance scoring, and distance-aware context pruning. Prerequisite OrcaLoca requires docker to run, so please first pull our docker image (forked from SWE-Agent): OrcaLoca also requires API access to LLM.
| Stars | 36 |
| Forks | 6 |
| Language | Python |
| Category | Agent Tool |
| License | MIT |
| Quality Score | 68.4467289722828/100 |
| Last Updated | 2025-04-07 |
| Created | 2024-06-21 |
| Platforms | python |
| Est. Tokens | ~460k |
Explore other popular agent tool tools:
OrcaLoca is OrcaLoca: An LLM Agent Framework for Software Issue Localization [ICML 25]. It is categorized as a Agent Tool with 36 GitHub stars.
OrcaLoca is primarily written in Python. It covers topics such as auto-code-refinement, llm, llm-agent.
You can find installation instructions and usage details in the OrcaLoca GitHub repository at github.com/fishmingyu/OrcaLoca. The project has 36 stars and 6 forks, indicating an active community.
OrcaLoca is released under the MIT license, making it free to use and modify according to the license terms.