Menu
Fetches AI news from Gmail Google Alerts; invoke via invoke_agent only. Returns date-prioritized list of title and url.
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
Decompose and solve complex multi-step problems. Creates a todo plan before executing, then works through each step — orchestrating subagents, memory, and tools. Inspired by LangChain plan-and-execute agents.
Analyze a user's codebase against a reference architecture to identify gaps and provide prioritized enhancement recommendations.
Coding workflow recipe (brainstorm → design → spec → plan → implement → review → verify → finish)
General-assistant workflow recipe (clarify → answer concisely → offer next step)
Ops/automation workflow recipe (assess → dry-run → execute → verify → report)
Research workflow recipe (frame question → gather sources → synthesize → cite → answer)
| name | news-fetcher |
| description | Fetches AI news from Gmail Google Alerts; invoke via invoke_agent only. Returns date-prioritized list of title and url. |
| tools | ["read_skill_file","mcp_google-workspace_query_gmail_emails"] |
| tags | ["news","gmail","subagent"] |
Single job: Query Gmail for Google 快訊 emails and return a list of news items with title and url, date-prioritized.
Use the Gmail MCP tool (mcp_google-workspace_query_gmail_emails) with query:
subject:Google 快訊 newer_than:2d
(Syntax may vary by MCP; use the equivalent that limits to last 2 days and subject containing "Google 快訊".)
Use each message's date to filter: first try today; if none, use yesterday.
For each matching email:
Return only a single block the main agent can parse. No prose before or after.
Use one of:
[{"title":"...","url":"..."},{"title":"...","url":"..."}]1. Title — URLNo commentary. No "Here are the results". Just the list.