Hi! The REFLECT and ARCHIVE modes generate documents (for example, reflection.md with lessons and the final project archive), but the system does not indicate how to use this information later. The Memory Bank lists reflect-*.md files as overview documents, but no mode or rule explicitly reads or takes their contents into account. As a result, valuable information recorded at the reflection/archive stages may be ignored by the AI in the future. The files are disposable: recorded, but do not affect subsequent steps.
Can you suggest a methodology for further use of files created in the REFLECT and ARCHIVE modes? Will this need to be added manually as context in the future? Or will the AI agent see this when indexing the code base and take it into account?
Hi! The REFLECT and ARCHIVE modes generate documents (for example, reflection.md with lessons and the final project archive), but the system does not indicate how to use this information later. The Memory Bank lists reflect-*.md files as overview documents, but no mode or rule explicitly reads or takes their contents into account. As a result, valuable information recorded at the reflection/archive stages may be ignored by the AI in the future. The files are disposable: recorded, but do not affect subsequent steps.
Can you suggest a methodology for further use of files created in the REFLECT and ARCHIVE modes? Will this need to be added manually as context in the future? Or will the AI agent see this when indexing the code base and take it into account?