solahqb
0 Followers · 0 Following
  • Joined on 2026-06-26
Repositories
2
 Projects Packages Public Activity Starred Repositories
solahqb pushed to main at solahqb/AgentEvalTool 2026-08-04 19:34:11 +00:00
da6ccc265d feat(frontend): restructure admin menu into two-level groups and add intelligent-eval entry
1317552701 feat(intelligent-eval): add backend for OpenClaw-driven intelligent evaluation (tickets 01-04)
Compare 2 commits »
solahqb pushed to main at solahqb/AgentEvalTool 2026-08-04 05:40:21 +00:00
f93d2f224f fix(frontend): auto-fetch report data when campaignId changes
solahqb pushed to main at solahqb/AgentEvalTool 2026-08-04 05:29:39 +00:00
160332665e refactor(exploration): absorb settlement.py into ExplorationSessionRepository
0aa3ef81c5 refactor(repository): extract AsyncJobRepository base class for analysis/comparison
cbfdf86b36 refactor(frontend): extract useCampaignReport hook from Campaigns.tsx
c24998c762 refactor(metrics): extract dashboard aggregation to compute_dashboard
42be31dd1f feat(report): add load_campaign_view as unified campaign read model
Compare 8 commits »
solahqb pushed to main at solahqb/AgentEvalTool 2026-08-03 19:47:47 +00:00
df76edcf55 refactor(exploration): move ledger and state machine into domain modules
38849d46f1 refactor(repository): narrow atomic updates for patrol/cancel/scheduler writes
f8d8450b1e refactor(report): unify campaign report loading behind one read model
Compare 3 commits »
solahqb pushed to main at solahqb/AgentEvalTool 2026-08-03 18:55:51 +00:00
a665b496b0 chore(v0.9): wrap over-length lines and record spec rulings
b26c432f3a refactor(frontend): reuse ChatBubble for exploration drill-down
ef4c094082 refactor(exploration): share one fetch+aggregate helper across outlets
936640fb36 fix(exploration): include all judge findings instead of poor-only
7f68afd765 docs(v0.9): mark ticket 06 done after Playwright browser verification
Compare 18 commits »
solahqb pushed to main at solahqb/AgentEvalTool 2026-08-03 10:24:16 +00:00
3e485bfbe6 docs(v0.9): mark ticket 03 patrol API done
6340ec503c feat(exploration): stateless patrol API with watermark increments
53afb9b5d1 feat(campaigns): exploration seed set and budget config per campaign
a351f65550 feat(exploration): session lifecycle endpoints with platform hard guardrails
c35159ffcb docs(v2): record exploratory-evaluation consensus in CONTEXT.md and ADR-0003
Compare 5 commits »
solahqb pushed to main at solahqb/AgentEvalTool 2026-08-03 07:24:23 +00:00
2d7679a6bb docs(release): freeze v0.8「比」period comparison milestone as 0.8.0
5ecb30876e style(tests): ruff 全量清理 — 49 项修复，backend 与 tests 全绿
14b09e1ac6 feat(campaigns): 周期对比纳入 Markdown 导出，活动导出排版重优化
dd3b9a5e91 refactor(comparison): 评审修复 — 共享 gateway_chat_client、指标元表、对比区块组件化
1e55a21649 feat(comparison): v0.8 周期对比 — 计划指纹自动基线配对、机械指标 diff 与 LLM 演进叙述
Compare 19 commits »
solahqb pushed to main at solahqb/AgentEvalTool 2026-07-31 07:11:51 +00:00
aa40c8e0d8 refactor(campaign): type Campaign.summary as CampaignSummary VO
ffc058f951 refactor(engine): thin _run_case and _save_rule_results
d411572607 refactor(storage): converge CRUD skeleton into BaseRepository
9c01afa79b refactor(engine): extract build_run_summary pure seam
983a58d013 refactor(verdict): unify read path on authoritative case_outcomes
Compare 5 commits »
solahqb pushed to main at solahqb/AgentEvalTool 2026-07-31 05:22:38 +00:00
050c674ee2 refactor(frontend): extract useResource/usePolling shared hooks
f285738f6d refactor(report): split report generation from pure rendering
0cca4963d1 refactor(tasks): unify run/campaign task registries into TaskRegistry
e815298ce5 refactor(campaign): move tick decisions into the pure scheduler seam
782916a283 refactor(metrics): type Run summary and converge cross-run aggregation
Compare 11 commits »
solahqb pushed to main at solahqb/AgentEvalTool 2026-07-30 03:45:23 +00:00
af7e7bf110 docs(v0.6): add Campaign spec and 5 tracer-bullet tickets
solahqb pushed to main at solahqb/AgentEvalTool 2026-07-30 03:13:07 +00:00
fe2a3f7579 docs(domain): add Campaign glossary terms and phasing ADR
solahqb pushed to main at solahqb/AgentEvalTool 2026-07-30 01:58:17 +00:00
1345daddd2 feat(engine): make poll_reply timeout configurable via env
solahqb pushed to main at solahqb/AgentEvalTool 2026-07-29 11:45:25 +00:00
5db0ede4f4 refactor(judgement): converge case-pass decision into one deep module
solahqb pushed to main at solahqb/AgentEvalTool 2026-07-29 07:43:35 +00:00
f1aa61edd0 chore: sync package-lock version to 0.5.0-dev
solahqb pushed to main at solahqb/AgentEvalTool 2026-07-29 07:43:24 +00:00
25b4c98dc8 docs(release): v0.5「准」发布说明与里程碑收尾
solahqb pushed to main at solahqb/AgentEvalTool 2026-07-29 07:35:42 +00:00
cbab55843b fix(reports): export via authed axios blob download instead of window.open
solahqb pushed to main at solahqb/AgentEvalTool 2026-07-29 07:26:47 +00:00
5dd3c4ff1c style(dashboard): redesign stat cards with fixed-height compact layout
solahqb pushed to main at solahqb/AgentEvalTool 2026-07-29 06:48:25 +00:00
d23b321225 fix(runs): mark orphaned running/pending runs failed on startup
solahqb pushed to main at solahqb/AgentEvalTool 2026-07-29 06:34:40 +00:00
fbac28bc7e feat(frontend): always show trigger source labels (手动/AI 助手/CLI)
solahqb pushed to main at solahqb/AgentEvalTool 2026-07-29 06:20:16 +00:00
8e3cf82fcc feat(frontend): surface scenario version across dropdowns and views