Skip to content

[feature] Recommendations: analyze token cost per context file and suggest modularization #948

Description

@ClaireGz

Problem

Users have no visibility into which context files consume the most tokens. Large monolithic files silently inflate LLM costs without the user knowing which files are the culprits.

Expected behavior

  • The recommendation bot analyzes token cost per context file
  • Flags the most expensive files with their estimated token count / cost contribution
  • Suggests modularization when a file is disproportionately large (e.g. "this file uses 40% of your context budget - consider splitting it into smaller focused files")
  • Surfaces these recommendations alongside existing context quality suggestions

Metadata

Metadata

Assignees

No one assigned

    Labels

    Type

    No type
    No fields configured for issues without a type.

    Projects

    Status
    TODO

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions