RSI TECH
Contact
Work index

open-source / lab / ongoing

Codebase Combiner

A workspace becomes a filtered, prompt-ready artifact with predictable size and token estimates.

System fingerprint

Select a signal node to inspect its operating principle.

  1. WorkspaceDiscover candidate source files.
  2. SelectionApply glob and extension filters.
  3. SizingEstimate prompt tokens before export.
  4. ArtifactGenerate one prompt-ready document.

01 / Evidence

Outcome

  • Two public deliverables for editor and desktop workflows.
  • Prompt-ready Markdown or text artifacts.
  • Filters and token estimates for context curation.

02 / Operating environment

Context

Workspace -> single Markdown/text prompt artifact with filters and token estimates (VS Code + macOS SwiftUI).

Developers preparing codebase context for LLM workflows

03 / Boundary conditions

Constraints

  1. C01

    LLM context needs the right files in the right order at a predictable size.

  2. C02

    The workflow must support both editor and desktop use.

  3. C03

    Prompt sizing must be visible before the artifact is used.

04 / Decision path

Architecture

Select a signal node to inspect its operating principle.

  1. WorkspaceDiscover candidate source files.
  2. SelectionApply glob and extension filters.
  3. SizingEstimate prompt tokens before export.
  4. ArtifactGenerate one prompt-ready document.

05 / Engineering choices

Key decisions

  • D1Ship both a VS Code extension and a macOS SwiftUI app.
  • D2Filter workspace content by glob and extension.
  • D3Generate a single Markdown or text artifact with token estimates.

06 / Proof discipline

Evaluation

  • Preview selected workspace content before generation.
  • Estimate tokens to make prompt size predictable.
  • Use filters to keep generated context scoped to the intended files.

07 / Scope

Ownership

VS Code extension, macOS SwiftUI application, filtering, and prompt artifact tooling

  1. O01

    VS Code extension core

  2. O02

    SwiftUI desktop application

  3. O03

    Filtering, preview, artifact generation, and token estimation

08 / Explicit compromises

Trade-offs

  • Prefer curated context over dumping an entire workspace.
  • Maintain two user interfaces around one prompt-packaging workflow.

Implementation stack

VS Code Extension / Node/JS / SwiftUI / Tooling

Next evidence

More work

delivered / Security AnalyticsSecurity Ops: User Behavior Analytics & Anomaly DetectionBehavioral baselines and explainable anomaly signals for SOC triage and investigation.delivered / Data PlatformsAutomated Time-Series Forecasting FrameworkA repeatable forecasting workflow made model comparisons safer and iteration easier to hand over.