chiguyong
|
3337589395
|
fix(review): document-processing code review fixes — validation, tests, formatting
Deploy to Production / deploy (push) Waiting to run
Details
- SkillConfig._validate_v2: validate fallback_strategies against
ReWOOEngine.VALID_STRATEGIES (lazy import, #20)
- test_skill_config: +4 tests for fallback_strategies validation
- test_document_loader: +8 xlsx edge case tests (empty workbook,
malformed bytes, column mismatch, row/cell truncation, multi-sheet,
file size limit, None cells, #16)
- test_execution_modes: fix ReWOOEngine patch path (lazy import ->
patch at source) + FakeReWOOEngine.execute return .output attribute
- config_driven: ruff formatting (quotes, blank lines after imports)
- project_rules: remove stale "known failing test" note (now passes)
|
2026-06-23 20:21:19 +08:00 |
|
chiguyong
|
bc424574c7
|
fix: Tauri reload, multi-conv blocking, skill install, UI polish
Deploy to Production / deploy (push) Waiting to run
Details
1) Tauri reload login: main.ts beginStartup before mount + router guard await waitForStartup
2) Multi-conversation blocking: per-conversation Map/Set tracking
3) Skill install: npx --yes + SKILL.md support + path traversal validation
4) Markdown table rendering + 80ms streaming debounce
5) Agent execution UI: structured IStreamingStep
6) auth.py _ensure_db idempotent
Code review fixes: renderTimer cleanup, counter accumulation, memory leak, WS reconnect stale steps
|
2026-06-23 11:03:46 +08:00 |