Menu
Browser-first news retrieval with paywall-aware stop policy, combining DuckDuckGo Lite discovery and RSS feeds.
Install with Codex or Claude Copy this prompt, paste it into Codex, Claude, or another assistant, and let it review the skill page and install it for you.
Based on SOC occupation classification
Slide layout library for PPT HTML slides - indexed layout contracts, selection rules, constraints, templates, and page budget guidance.
Detect document domain (14 domains: software/hardware/manufacturing/standards/biotech/business/energy/data_science/automotive/cloud_infrastructure/telecom/iot/medical_devices/security) via keyword matching. Supports multi-domain documents and confidence scoring.
Automated validation, reporting, and fixing for Markdown documents using markdownlint, prettier and md_table_tool.
Detect and fix Markdown table alignment issues (MD060) with display-width awareness for CJK characters.
Persistent memory management for AI agents - 3-tier file-based memory system with mandatory L1 raw logs and optional L2/L3 extraction.
Brand-style design system for PPT HTML slides - single source of truth for 6 reusable style profiles, including consulting-style and editorial-briefing styles.
| name | news-search |
| description | Browser-first news retrieval with paywall-aware stop policy, combining DuckDuckGo Lite discovery and RSS feeds. |
| metadata | {"version":"3.0.0","author":"cortana"} |
This skill retrieves news with a browser-first workflow to reduce brittle crawler behavior.
Flow:
duckduckgo and/or rss# Browser-first (default), all sources
python3 skills/news-search/scripts/search.py "OpenAI latest model" --source all --max 5
# RSS-only discovery + browser retrieval
python3 skills/news-search/scripts/search.py "AI" --source rss --category ai --max 5
# HTTP fallback mode
python3 skills/news-search/scripts/search.py "global markets" --retrieval http --max 5
# Skip detail retrieval (discovery only)
python3 skills/news-search/scripts/search.py "chip industry" --retrieval none --max 8
# Continue even if paywall markers are found
python3 skills/news-search/scripts/search.py "financial times" --allow-paywalled
--source: duckduckgo | rss | all (default: all)--category: tech | finance | sports | entertainment | global | domestic | military | gaming | ai--max: maximum results (default: 5)--retrieval: browser | http | none (default: browser)--allow-paywalled: disable stop-on-paywall behavior--timeout: request timeout in secondsThe script returns JSON with:
results[*].retrieval_mode: browser | http | noneresults[*].access: open | paywalled | unknownresults[*].paywall_reason: matched signal when paywalledmetadata.paywalled_countmetadata.warnings (for fallback events like missing Playwright)Default policy is strict:
access=paywalledhttp retrieval and reports a warning.