Skip to main content

Feature Request: AI Dev Kit Project Review and Legacy Clean-Up

Type: Feature Request (FR)
Submitted: 2026-02-19
Submitted By: User
Priority: HIGH
Status: IN PROGRESS
Assigned Task: E5:S01:T39
GitHub Issue: [TBD]

Summary​

Perform a full AI Dev Kit repository review to identify and then eliminate or mitigate legacy, redundant, and confusing artefacts (docs, scripts, configs, examples) so that ai-dev-kit is a clean, coherent reference implementation for adopters and for the accompanying book.\n+​

Description​

What is the Feature?​

Establish a systematic, documented review of the ai-dev-kit repository that:

  1. Audits framework packages, dev-kit-local docs, Kanban, workflows, and scripts.
  2. Clearly distinguishes framework sources of truth from dev-kit-specific specialisations.
  3. Identifies legacy, duplicated, or confusing artefacts and produces a concrete clean-up plan (what to keep, consolidate, archive, or remove).

This FR is review-only; actual deletions/moves/edits will be performed under a follow-up implementation task once the review report and clean-up checklist are approved.

What Problem Does This Solve?​

  • Over time, ai-dev-kit has accumulated:

    • Historic copies of policies that were later extracted into framework packages.
    • Multiple RW/UKW/CMW docs representing earlier iterations.
    • Legacy or experimental scripts alongside canonical workflow tooling.

  • For adopters and for the book, we want ai-dev-kit to look like a sharp, minimal reference:

    • Clear canonical docs.
    • Minimal duplication.
    • Obvious paths for installing and using frameworks.

This FR ensures we understand and document what should change before any structural clean-up is done.

What is the Use Case?​

  • Adopters: Projects (like Starborn Legacy, Confidentia, fynd.deals) using ai-dev-kit as a reference or submodule should see one coherent set of docs and tools.
  • Book project: We want to be able to say “this repo is the canonical example” without tripping over historic artefacts in every chapter.
  • Framework evolution: Future changes to frameworks should be easier to propagate when the current layout is clean and clearly documented.

Requirements​

Functional Requirements​

  • FR-039:R01 – Create a top-level inventory of major ai-dev-kit directories, classifying them as:

    • Framework SoT.
    • Dev-kit-local specialisation.
    • Likely legacy or transitional.

  • FR-039:R02 – Review each core framework package (numbering & versioning, workflow mgt, kanban, etc.) for:

    • Legacy or superseded docs.
    • Redundant examples or partial drafts.

  • FR-039:R03 – Identify and document duplication between framework docs/scripts and dev-kit-local copies (policies, RW docs, Kanban governance, etc.).

  • FR-039:R04 – Review Kanban and workflow documentation (RW/UKW/CMW) for:

    • Outdated patterns.
    • Conflicting narratives or overlapping guides.

  • FR-039:R05 – Review versioning and changelog artefacts (version file, rw-config.yaml, CHANGELOG.md, archive) for:

    • Consistency with the current dev-kit versioning policy.
    • Stray or obsolete files.

  • FR-039:R06 – Catalogue scripts and helpers, tagging each as:

    • Core framework/RW/UKW tooling.
    • Project utility.
    • Legacy/experimental.

  • FR-039:R07 – Produce a single markdown review report summarising findings and proposing:

    • What to keep as canonical.
    • What to consolidate or merge.
    • What to archive as historical.
    • What to remove.

  • FR-039:R08 – Derive from the report a sequenced clean-up checklist suitable for a follow-up implementation task.

Non-Functional Requirements​

  • FR-039:NF01 – The review itself is non-destructive: it must not delete/move files; it only reads and reports.
  • FR-039:NF02 – The report must be clear enough that an implementation task can follow it without re-discovering context.
  • FR-039:NF03 – The review should favour clarity over perfection: it’s acceptable to mark some items as “requires human judgement” in the clean-up checklist.

Scope Analysis​

Problem Domain: Repository hygiene, framework/documentation structure, adoptability
Affected Areas:

  • packages/frameworks/*
  • docs/architecture/standards-and-adrs/*
  • docs/project-management/kanban/*
  • src/fynd_deals/version.py, rw-config.yaml
  • CHANGELOG.md and docs/changelog-and-release-notes/changelog-archive/*
  • scripts/*

Estimated Complexity: Medium (wide read-only survey + one coherent report).\n+​

Acceptance Criteria​

  • Inventory of major directories with classification (framework SoT / dev-kit-local / likely legacy).
  • Per-framework notes on legacy/superseded docs/scripts.
  • Documented list of duplicated policies/docs/scripts and suggested consolidation targets.
  • Summary of Kanban and workflow docs, with outdated/conflicting items flagged.
  • Versioning/changelog consistency check with any stray files listed.
  • Script/helper catalogue with legacy/experimental items flagged.
  • A markdown review report and ordered clean-up checklist exist and are referenced from this FR and its Task.

Dependencies​

Blocks:

  • Confidently using ai-dev-kit as the canonical reference for adopters and for the book.

Blocked By:

  • None (this is review-only; implementation clean-up work will be a separate task).

Related Work:

  • Framework policies and implementation guides under packages/frameworks/*.
  • RW/UKW/CMW agent execution guides.
  • FR Repo tasks for prior policy and workflow hardening.

Release History​

  • v0.5.1.39+3 (2026-02-19) – Legacy clean-up implementation: SoT references in dev-kit policies, scripts moved to scripts/internal/, adopter-facing repo structure doc.
  • v0.5.1.39+2 (2026-02-19) – Latest-only changelog policy and CMW mode; CHANGELOG.md keeps only Unreleased + newest release.
  • v0.5.1.39+1 (2026-02-19) – Initial review release for AI Dev Kit project review and legacy clean-up (planning and review docs only).