W25 Snapshot — As of 2026-06-15 06:50 SGT
League Table — Sprint 4 Final · W24
| Team | Level | Grade | Tags (correct?) | CI | Workflows | Branches | PRs | Contributors | FigJam Filled | FigJam Authors | Language |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Team 1 — sinder38 | L3+ | A | vW22 · vW23 · vW24 ✅ | ✅ PASSING | 3 active | 20 | 10 | 11/11 | 100+ (too large) | 11 | HTML + Python |
| Team 2 — LuJunwen212 | L2-3 | B+ | vW22 · vW23 · vW24 ✅ | ✅ PASSING | 1 active (×5 runs) | 5 | 9 | 11/11 | 137 / 401 (34%) | 2 | Python |
| Team 3 — juninho4 | L1-2 | B | vW22 · vW23 · vW24 ✅ | ⬜ NONE | 0 workflows | 1 | 0 | 11/11 | 96 / 410 (23%) | 3 | Jupyter Notebook |
| Team 4 — wintwah-lwin | L1-2 | C+ | W23 · W24 · v1.0.0 ⚠️ Wrong format | ⬜ NONE | 0 workflows | 1 | 0 | 8/11 | 85 / 360 (24%) | 4 | TypeScript |
| Team 5 — Gong-yinxuan | L2 | B- | W23 · vW24 ⚠️ Mixed format | ✅ PASSING | 1 active (fragile) | 4 | 0 | 8/11 | 23 / 349 (7%) | 1 | HTML |
Team 1 — sinder38/Team-1-Prac-A-Project · EXCEPTIONAL
Team 1 — sinder38
- 2026-06-15Create prediction_2026-W24_Team1.md (#55)
- 2026-06-14chore: pipeline artifacts 2026-06-14 [bot PR merged]
- 2026-06-14feat(llm): automate multi-model synthesis pipeline with OpenRouter API (Level 3)
- 2026-06-14feat: automated pipeline runner with CI/CD workflow
- 2026-06-14feat/fix: refactor LLM pipeline for strict fail-fast execution and robust parsing
- 2026-06-14feat: Almanac Agent Week 4 analysis complete
- 2026-06-14feat(macro): W24 macro agent output committed
- 2026-06-14feat(technical): W24 technical agent signals
- 2026-06-13feat(r10): delta scoring engine v2
- 2026-06-07chore: sprint 3 retrospective and W23 tag
ge4Qba5sqTe1MvN8sVH4d4A fully automated, production-grade market intelligence pipeline. The system runs itself: GitHub Actions triggers on schedule → fetches market data → runs agents (Almanac, Macro, Technical) → calls OpenRouter API to synthesise ChatGPT + Gemini + DeepSeek → opens a Pull Request with the artifacts → human score applied → prediction sealed as a tagged release. This is not student work — this is junior developer work.
No intervention needed. They will deliver vW25 ahead of deadline. The only question for Sprint 5 is whether they improve prediction accuracy (delta scoring) and whether the Wild Card Human Score shows independent analytical thinking beyond the LLM consensus.
The role-branching model (20 branches, each named after a role/week) is the reference architecture for future cohorts. The automated PR bot is the key differentiator — no other team has this. sinder38's commit rate (47 commits) is 5× the next individual contributor on any team.
sinder38 carries a disproportionate commit load (47 vs next at 9). This is excellent individually but the course requires role contribution from all 11 members. Confirm R3, R4, R5, R7, R10 can speak to their own outputs on Monday without sinder38 doing it for them.
Team 2 — LuJunwen212/CP3405-Project · SOLID
Team 2 — LuJunwen212
r6_macro_automation.py — Python macro data fetcherprediction_2026-W22/W23/W24_team2.md — predictions filed every sprintactual_2026-W22.md · actuals_2026-W23.md — actuals trackedmarket_snapshot_W24.json — automated JSON data captureDECISION_2026-W24.md — decision log maintained
- 2026-06-14Update prediction_2026-W24_team2.md
- 2026-06-14Update market_snapshot_W24.json [automated]
- 2026-06-14feat: r6 macro automation script v2
- 2026-06-14Add LLM synthesis W24 — ChatGPT/Gemini/DeepSeek outputs
- 2026-06-14Add DECISION_2026-W24.md
- 2026-06-14Merge Week-04 branch into main
- 2026-06-14Add actuals W24 — SPX NDX IWM closes
Repeated "Upload files / Delete / Upload files" pattern detected — several members uploading through GitHub's browser interface rather than using Git CLI. This is manual file management disguised as commits. The pipeline itself is automated but individual contribution method is not.
13ljuKDXj13gnoknGb4EcH✅ Values — filled (Commitment, Focus, Openness, Respect, Courage)
✅ Member cards — Zheng Zaikun, Wu Qinyang, Kan Yijie, Guanyu Lu identified
✅ Sprint backlog — Sprints 1–5 sections present
✅ Communication / Accountability / Conflict Resolution sections
⚠️ Role assignments — "Product Designer", "Product Marketing Manager" still showing as placeholders
❌ 264 stickies empty — 66% of board unfilled
11/11 contributors all active. Automated pipeline ran 5 green CI runs on sprint deadline day — this shows discipline. market_snapshot_W24.json auto-captured. Actuals tracked across W22, W23, W24. Decision log maintained. Second most complete repo after Team 1.
Only 1 branch (Week-04) — no role-based branching. The Bao Le is carrying most of the PR merging. For Sprint 5 the Scrum Master must implement proper branch-per-role workflow. FigJam participation from 9 missing members is the other urgent gap.
Team 3 — juninho4/Team-3-Practical-A-CP3405 · ACTIVE BUT MANUAL
Team 3 — juninho4
vW22/ · vW23/ · vW24/ — sprint-by-sprint foldersactuals_2026-W24.md — R6 filing actuals correctlylearning_log_2026_vW24.md — reflective learning documentedacceptance_criteria.md — structured thinking visible*.ipynb — Jupyter Notebooks with real data analysisRepo is 40MB — suggests substantial chart/screenshot evidence files
- 2026-06-14Add files via upload — vW24 evidence
- 2026-06-14Upload market analysis notebook W24
- 2026-06-14Add actuals_2026-W24.md
- 2026-06-14Add LLM synthesis outputs W24
- 2026-06-14Add technical agent analysis W24
- 2026-06-14Add learning_log_2026_vW24.md
- 2026-06-14Cut vW24 release tag
HBSqbX0gYOL9rzhhU40Emi✅ Member cards — Juninho Chandra, Chadric Kok, Jiahai Xiong, Wang Zhiyuan, Asher Mulligan
✅ "I often need to fully understand what I need to do before I start" — candid, genuine
❌ Vision/Mission sticky exists but is blank
❌ No vW tag references anywhere on board
❌ 314 stickies empty — 77% unfilled
⚠️ Previously audited: some stickies were copied — now shows some original content
Largest repo (40MB) and highest total commit count across all 5 teams. 11/11 members visible on GitHub — the most democratically contributing team. Jupyter Notebooks suggest real data analysis is happening, not just file uploads. Folder structure (vW22/vW23/vW24) is clean.
Zero GitHub Actions workflows. Single branch. Zero pull requests. All commits are browser uploads ("Add files via upload"). The team is doing real work but not learning Git properly. Sprint 5 single instruction: add one GitHub Actions workflow. That is all.
Team 4 — wintwah-lwin/CP3405_Group_4 · SURPRISING
Team 4 — wintwah-lwin
vW23 · vW24 — missing the "v" prefix. Correct immediately for Sprint 5. v1.0.0 is completely off-format.This is a proper named market intelligence dashboard application — not just a script.
DEPLOYMENT.md — existed (deployment guide was written)API_GUIDE.md — existed (API documentation was written)⚠️ TODAY (2026-06-15): Arisa is deleting DEPLOYMENT.md and API_GUIDE.md — suggests the dashboard was over-scoped and being trimmed back before Monday's presentation.
evidence/ folder — Finviz screenshots committed
- 2026-06-15Delete DEPLOYMENT.md [scaling back TradeKyaMal]
- 2026-06-15Delete API_GUIDE.md [scaling back TradeKyaMal]
- 2026-06-14Add prediction W24 — SPX NDX direction
- 2026-06-14Add macro agent W24 output
- 2026-06-14Add technical analysis W24
- 2026-06-13Cut W24 tag [note: wrong format — should be vW24]
- 2026-06-07feat: TradeKyaMal TypeScript dashboard initial commit
5E8UzMZyBtcaRIUeDCLAZ0"To become a leading team in market intelligence analysis..."
Full Mission covering multi-agent analysis, AI synthesis, human scoring
✅ vW22 tag referenced on board — shows tag awareness
✅ Values — transparency, ownership, collaboration (filled)
⚠️ Member cards — several still showing "Your Name / Your Title" placeholders
❌ 275 stickies empty — 76% unfilled
thihanhtun wrote the strongest Vision/Mission statement of any team on FigJam. TradeKyaMal shows genuine ambition — someone in this team wanted to build a real product. 4 FigJam authors is the best of teams 2-5.
Tag format is wrong — W24 must become vW24. This will affect delta scoring. wintwah-lwin must fix this immediately. Also 3 contributors not visible on GitHub at all — these members need to be accounted for.
Deleting DEPLOYMENT.md and API_GUIDE.md today suggests the team overbuilt and is now scaling back. Ask them to demo whatever TradeKyaMal does — even partially. A TypeScript dashboard is worth acknowledging regardless of completion status.
Team 5 — Gong-yinxuan/CP3405 · BETTER THAN FIGJAM SUGGESTS
Team 5 — Gong-yinxuan
Bot account:
HYH-design-botThis shows real initiative — a student who felt "out of place" on FigJam built an actual bot.
19 automated commits is more than most human contributors on other teams.
human_score_W04.md — R7 Human Score filed ✅llm_synthesis.md — LLM synthesis documented ✅decision.md — decision log maintained ✅prism/ folder — possibly referencing PRISM platform
- 2026-06-14Update llm_synthesis.md — W24 ChatGPT/Gemini/DeepSeek outputs
- 2026-06-14Add human_score_W04.md — R7 Wild Card applied
- 2026-06-14[bot] Auto-update market data 2026-06-14
- 2026-06-13update AI_prompt — refine LLM synthesis prompt
- 2026-06-13[bot] Scheduled market data update 05:18 SGT
- 2026-06-13Cut vW24 tag
- 2026-06-13Add decision.md W24
gwwJXhPuxuZ1Ohwzai2kKaHo's profile note: "I often feel out of place skipping 2 modules" — onboarding incomplete. Yet Ho built HYH-design-bot with 19 commits.
❌ Values — empty
✅ Ho Yuet Hei member card — R8 Evidence Collector, Sprint 1
⚠️ Only "Good collaboration" and template rows filled
❌ No vW tag references on board at all
❌ Sprint mentions go back to "Week 01" only — no progression visible
FigJam score of 7% would suggest this team is nearly inactive. GitHub tells the opposite story. ToHoangGia and Gong-yinxuan are both at 48 commits each — the most balanced top-two contribution of any team. HYH-design-bot with 19 automated commits shows genuine automation initiative. llm_synthesis.md, human_score, and decision.md all filed.
Ho felt "out of place" on FigJam yet built an automation bot. This is the kind of student who needs a public acknowledgement to stay engaged. Call out HYH-design-bot specifically.
Only 8 contributor accounts visible. 3 members have zero GitHub commits and zero FigJam stickies. These students need to be identified by name and asked to show their work on Monday.