Private

Public Access

Files

T

ed 102f219904 docs(artifacts): Phase 2 wrapper inventory (9 P1 cruft sites; per-file mapping for Phases 3-7)

Phase 2 inventory output: 9 legacy wrappers (all P1 drop-errors-via-.data).
- Phase 3 (mcp_client): 1 (_resolve_and_check)
- Phase 4 (ai_client): 5 (_reread_file_items, _list_anthropic_models, _list_gemini_models, _extract_gemini_thoughts, _list_minimax_models)
- Phase 5 (rag_engine): 1 (_chunk_code)
- Phase 6 (gui_2): 2 (_detect_refresh_rate_win32, _resolve_font_path)

Source-of-truth note: PHASE1_AUDIT_BASELINE.json was gitignored and lost;
this inventory was regenerated from a current-tree scan via
scripts/audit_legacy_wrappers.py (revised to exclude the proper _result
helpers themselves from the wrapper pattern).

2026-06-20 19:41:48 -04:00

3.4 KiB

Raw Blame History

Phase 2 — Legacy Wrapper Inventory

Generated: 2026-06-20 Total legacy wrappers found: 9 (all P1 — return _x_result(...).data patterns)

Source-of-truth note

The tests/artifacts/PHASE1_AUDIT_BASELINE.json was lost (gitignored; not in any branch). This inventory was generated by scripts/audit_legacy_wrappers.py against the CURRENT (post-migration) src/ tree, which is the correct surface for the wrapper removal work.

The audit script (revised in this phase) flags ONLY functions that:

Have a _x name (private)
Their body calls _x_result(...) (the proper Result helper)
They unwrap .data (P1) or just return the Result (P3)

Functions whose own name ends in _result (the proper Result helpers themselves) are NOT wrappers — they ARE the Result-returning helpers.

Per-file legacy wrappers (9 total)

#	File	Line	Wrapper	Pattern	In-site callers	Notes
1	src/mcp_client.py	175	`_resolve_and_check`	P1	7 (internal use + dispatch_tool_call)	Returns tuple[Path\|None, str]; drops ErrorInfo structure. Used as a quick-rejection helper inside the MCP dispatch boundary.
2	src/rag_engine.py	250	`_chunk_code`	P1	2 (self + caller)	Returns str on success, formatted error on failure; drops ErrorInfo.
3	src/gui_2.py	227	`_detect_refresh_rate_win32`	P1	2 (self + caller)	Returns float refresh rate; on failure returns 0.0 (drops ErrorInfo).
4	src/gui_2.py	277	`_resolve_font_path`	P1	2 (self + caller)	Returns Path or None on failure (drops ErrorInfo).
5	src/ai_client.py	1073	`_reread_file_items`	P1	5	Returns list[dict]; drops ErrorInfo.
6	src/ai_client.py	1358	`_list_anthropic_models`	P1	2	Returns list[str]; drops ErrorInfo.
7	src/ai_client.py	1668	`_list_gemini_models`	P1	2	Returns list[str]; drops ErrorInfo.
8	src/ai_client.py	1815	`_extract_gemini_thoughts`	P1	2	Returns str; drops ErrorInfo.
9	src/ai_client.py	2487	`_list_minimax_models`	P1	5	Returns list[str]; drops ErrorInfo.

Patterns NOT flagged (and why)

All def _x_result(...): functions (91 total): these ARE the proper Result helpers.
Public def tool_name(...): functions in mcp_client.py (search_files, list_directory, etc.): These are MCP protocol boundaries; they MUST return str per the JSON-RPC wire format. They are protocol drain points, not internal cruft.
Functions that don't call their sibling _result variant: not wrappers.

Audit script revision

The original scripts/audit_legacy_wrappers.py (per spec §Plan Task 2.1) had a bug: it flagged any function with _result( in its body, including the proper helpers themselves (which call OTHER _result helpers). The fix is in is_legacy_wrapper: require the function name NOT to end in _result, AND the body must call name + "_result" specifically (its sibling helper).

This narrowed the finding from 111 (false-positive inflated) to 9 (true legacy wrappers).

Phase 3-7 mapping

Phase	File	Wrappers
3	src/mcp_client.py	`_resolve_and_check` (1)
4	src/ai_client.py	`_reread_file_items`, `_list_anthropic_models`, `_list_gemini_models`, `_extract_gemini_thoughts`, `_list_minimax_models` (5)
5	src/rag_engine.py	`_chunk_code` (1)
6	src/gui_2.py	`_detect_refresh_rate_win32`, `_resolve_font_path` (2)
7	(none)	—

3.4 KiB Raw Blame History