Vietnam Market News Monitor (Web Workflow)

Use this skill to monitor domestic Vietnam market narratives by sector and ticker mentions.

Data sources

• CafeF: https://cafef.vn/
• VnEconomy: https://vneconomy.vn/
• VietnamFinance: https://vietnamfinance.vn/
• Vietstock: https://vietstock.vn/

Tooling assumption

• OpenClaw web fetch + Brave API available.
• No Python scripts needed for this skill.

Execution workflow

• Fetch latest headlines from all four sources.
• Tag headlines by sector and potential ticker mentions.
• Count mention intensity by sector.
• Compare with previous snapshot (if available) to detect narrative acceleration/deceleration.
• Highlight possible misinformation/noise risk areas.

Data quality gate (required)

Run this gate before final output:

• Freshness: include collection window in ICT (UTC+7) and last-update timestamp per source.
• Coverage: target >= 3 domestic sources; if < 3, set confidence to Low.
• De-duplication: cluster duplicate headlines by normalized title + publisher domain.
• Evidence tags: mark each major claim as Fact (headline/text) or Inference (market implication).
• Bias check: flag concentration if one source contributes > 50% of collected items.

Shared confidence rubric (required)

Apply this common standard:

• High: >= 3 active sources, fresh window (24h/7d explicitly stated), and low concentration bias.
• Medium: 2 sources or moderate concentration bias, but sector/ticker narratives are still cross-validated.
• Low: < 2 sources, stale window, or highly concentrated/noisy coverage with weak confirmation.

Always output confidence with:

• Source coverage table.
• Freshness window (ICT UTC+7).
• Major noise/bias factor that may distort narrative ranking.

Output format (required)

• Section 1: Source coverage and freshness.
• Section 2: Top sectors by media intensity.
• Section 3: Most-mentioned tickers.
• Section 4: Narrative shift vs previous period.
• Section 5: Risks and validation checklist before investment action.

Watchlist mode (optional)

If the user provides an ACTIVE_WATCHLIST (tickers), add an extra section:

• Watchlist Impact Map: for each ticker, list relevant headlines (with links), label Fact vs Inference, and provide 1–2 monitoring triggers.

Do not issue buy/sell commands; this skill is for signal monitoring only.

Trigger examples

• "What are the hottest Vietnam market narratives today?"
• "Which sectors and tickers are mentioned most in Vietnamese financial media?"
• "Compare this week's market narrative vs last week."

Quality rules

• Headlines are signals, not direct trading decisions.
• Always recommend cross-checking with vnstock quantitative metrics.
• Provide source links for every major statement.
• If source coverage is weak or stale, explicitly downgrade confidence.