Increment mode quits with ValueError: call with_srcs first

[BakedJesus]

Bug description Increment mode quits with ValueError: call with_srcs first

Bug solved method Python script: if __name__ == "__main__": repo_existing_path = "./crypto_dash" repo: ProjectRepo = generate_repo(idea = "Convert this repo so that it is object-oriented style architecture, complete with beautiful dashboard with many elements", n_round=50, run_tests=False, inc=True, project_path=repo_existing_path) print(repo)

Output: Traceback (most recent call last): File "/home/qavi/repos/meta_generator.py", line 13, in <module> print(repo) File "/home/qavi/penvs/metaenv/lib/python3.10/site-packages/metagpt/utils/project_repo.py", line 107, in __str__ srcs_str = f"Srcs({self.srcs.all_files})" File "/home/qavi/penvs/metaenv/lib/python3.10/site-packages/metagpt/utils/project_repo.py", line 125, in srcs raise ValueError("Call with_srcs first.") ValueError: Call with_srcs first.

Environment information

• LLM type and model name: GPT-4 turbo-preview
• System version: Ubuntu 23.04
• Python version: conda python 3.10.13

• packages version:
• installation method: pip install metagpt

Screenshots or logs

[iorisa]

--project-path points to an old version project repository generated by MetaGPT. Importing an external third-party repository requires a reverse-engineer module, which has not been released yet. More details: https://docs.deepwisdom.ai/main/en/guide/in_depth_guides/incremental_development.html

[geekan]

@BakedJesus Did this answer solve your problem? It means that you need to use the new version to generate and propose incremental requirements