Add research-radar skill for autonomous research briefings - #468
Add research-radar skill for autonomous research briefings#468AtakanGs wants to merge 3 commits into
Conversation
|
Tested manually on Windows (WSL Ubuntu). How to test: Start Hermes Invoke the research-radar skill Generate a briefing for a topic like AI coding agents Confirm the Markdown output is written successfully Confirm dated output filenames are created as expected Example output: ai_agents_briefing.md ai_agents_briefing_2026-03-06.md |
|
This skill was demonstrated in a working Hermes setup (WSL Ubuntu). Demo workflow:
Example output files:
|
|
Hey @AtakanGs, thanks for the submission and for participating in the hackathon! 🎉 We're not going to merge this as a bundled skill — bundled skills ship with every Hermes install and need to be broadly useful to most users. This one is more of a prompt template (no code/scripts, and it artificially restricts the agent from using That said, this would work great as a community hub skill! Hermes has a Skills Hub system designed exactly for this. Here's how to publish it: Option 1: Publish to a GitHub repo (recommended) # Create a GitHub repo for your skills (e.g. AtakanGs/hermes-skills)
# Then publish directly from the CLI:
hermes skills publish skills/research/research-radar --to github --repo AtakanGs/hermes-skillsOther users can then install it with: hermes skills tap add AtakanGs/hermes-skills
hermes skills install research-radarOption 2: Local skill Tips to make it better:
Check out the docs on publishing skills: https://hermes.nousresearch.com/docs/developer-guide/creating-skills Thanks again for contributing! |
This PR introduces a new Hermes skill called
research-radar.The skill monitors topics across public sources, extracts high-signal developments, and generates structured research briefings in Markdown.
Core skill features:
What Changed Since Last RunExample use case:
Generate a recurring briefing about a topic such as AI coding agents and save it as a dated Markdown report.
This skill was developed as part of the Hermes Agent Hackathon.
How to test:
research-radarskillWhat Changed Since Last RunExample output files:
ai_agents_briefing.mdai_agents_briefing_YYYY-MM-DD.mdai_agents_briefing_YYYY-MM-DD_v2.mdai_agents_briefing_YYYY-MM-DD_v3.mdFurther product/system layer built on top of the skill:
This demonstrates that
research-radarcan function not only as a reusable Hermes skill, but also as the backbone of an interactive multi-topic intelligence workflow.