topic-agent_skill

This skill orchestrates today's topic workflow by collecting AI hotspots, generating proposals, auditing quality, and saving results to the Obsidian topic
  • Python

79

GitHub Stars

1

Bundled Files

2 months ago

Catalog Refreshed

4 months ago

First Indexed

Readme & install

Copy the install command, review bundled files from the catalogue, and read any extended description pulled from the listing source.

Installation

Preview and clipboard use veilstrat where the catalogue uses aiagentskills.

npx veilstrat add skill zephyrwang6/myskill --skill topic-agent

  • SKILL.md6.2 KB

Overview

This skill is the topic-agent: a topic-selection controller that coordinates hotspot collection, idea generation, and review until approved topics are produced and stored in an Obsidian topic library. It supports full daily runs, hotspot-only collection, single-topic analysis, and fast recommendations. Outputs are saved as markdown files ready for writing and tracking.

How this skill works

The agent collects daily hotspots across five categories (blogger practices, new products, vendor updates, research, community discussion) via web search, extracts actionable signals, and generates a shortlist of topic proposals with event description, angle, title options, writing style, outline, and source links. Each proposal is scored on heat, uniqueness, and domestic relevance; high-scoring items are saved to the Obsidian topic vault, while lower scores receive concrete improvement suggestions and can iterate until acceptable.

When to use it

  • Start a full daily workflow: collect hotspots → generate → review → save
  • Quick snapshot of today's AI hotspots without creating topics
  • Analyze and validate a single topic you already have
  • Get 3–5 recommended high-potential topics based on recent trends
  • Continue or modify an interrupted workflow or adjust angles

Best practices

  • Prefer primary-source links for every hotspot entry (specific article or post URL)
  • Prioritize content with practical, repeatable value for readers
  • Use the three-dimension scoring (heat, uniqueness, domestic interest) to filter focus
  • Iterate on angles for borderline topics rather than discarding immediately
  • Name and save approved topics with clear metadata and source links for traceability

Example use cases

  • Morning editorial run: produce 10 topic proposals and save 3–5 approved ones into Obsidian
  • Fetch today’s AI product launches and blog tactics to inform a hands-on tutorial
  • Validate and refine a reader-submitted idea into a publishable topic with scoring and improvement steps
  • Quickly recommend evergreen or timely topics for social posts or newsletter pitch
  • Audit a candidate topic’s heat by searching community threads and recent vendor announcements

FAQ

Hotspots are collected via parallel web searches across five categories—blogger practice, new products, vendor updates, research, and community discussion—and organized with direct article/post URLs and concise summaries.

What determines whether a topic is saved?

Topics are scored on heat (30%), unique angle (40%), and domestic relevance (30%). Scores above 80/100 are saved; 60–79 receive revision advice; below 60 are rejected with reasons.

Built by
VeilStrat
AI signals for GTM teams
© 2026 VeilStrat. All rights reserved.All systems operational
topic-agent skill by zephyrwang6/myskill | VeilStrat