Share one thing about Hermes that you are pretty sure most users don’t know about! I’ll start - and will commit to compiling these all as a post next
/Handoff - https://hermes-agent.nousresearch.com/docs/user-guide/sessions#cross-platform-handoff
@Teknium @max_paperclips @NousResearch
Many users praise Hermes Agent features like cross-platform session handoffs and /claude-code as super handy and the GOAT because they enhance usability and workflows.
No Digg Deeper questions have been answered for this story yet.
@hermes_updates @max_paperclips @NousResearch /browser connect
https://hermes-agent.nousresearch.com/docs/user-guide/features/browser#local-chromium-family-browser-via-cdp-browser-connect
@hermes_updates @Teknium @max_paperclips @NousResearch Hermes has its own api https://hermes-agent.nousresearch.com/docs/user-guide/features/web-dashboard#rest-api
@hermes_updates @Teknium @max_paperclips @NousResearch /claude-code. Honestly the boon that Anthropic left -p available has been great. Hermes can ask Claude to run whole skill workflows, so if you had Claude set up it’s really just about adding it to the fleet. Super useful skill.
@hermes_updates @Teknium @max_paperclips @NousResearch It’s new enough I’ll even go out on a limb and say the native cross platform Hermes desktop app
https://hermes-agent.nousresearch.com/docs/user-guide/desktop
Hey @witcheer, really useful breakdown on the compression behavior.
When a detail from the middle of a long session gets summarized away, how often do you see the agent repeat work it already did?
In those cases, how much do you recover through external memory + structured context (like ALIVE) versus tuning the three levers in config.yaml? Have you found any reliable patterns to protect high-value intermediate decisions or tool outputs during summarization?
@hermes_updates @Teknium @max_paperclips @NousResearch I can’t believe I didn’t know about this!!
@hermes_updates @Teknium @max_paperclips @NousResearch using /steer to redirect after you already entered a prompt and it’s working
@hermes_updates @Teknium @max_paperclips @NousResearch I need this I'm new to Hermes
Good one!
Compression pro tips
@rouk @Teknium @max_paperclips @NousResearch /steer is the GOAT! https://hermes-agent.nousresearch.com/docs/user-guide/cli#busy-input-mode
@nesquena @Teknium @max_paperclips @NousResearch If you all haven’t tried the desktop app, definitely take a look!
@vigilharbor @Teknium @max_paperclips @NousResearch Great one! Love /claude-code
@Teknium @max_paperclips @NousResearch Ooh good one! Thanks Tek!
@MoonEmpirE0 @Teknium @max_paperclips @NousResearch Welcome!
@fpostigo @Teknium @max_paperclips @NousResearch Yeah it’s super handy, give it a try!
@tom_mulkins @Teknium @max_paperclips @NousResearch Great one!
Hermes Wingtips #2
the 🗜️ icon, what context compression keeps and what it drops
this little emojy is the compression count: how many times Hermes auto-summarised the session to stay under the context limit, which kicks in around 50% full.
when compression occurs, it keeps your first 3 turns and your last 20, and summarises everything in between. a detail from the middle of a long session can then drop out, and the agent repeats work it already did, even though the opening goal and the recent turns are intact.
~~~ three levers when it bites, all in config.yaml, hot-reloading on a running gateway:
- protect_last_n keeps more recent turns uncompressed
- auxiliary.compression.model points the summariser at a cheap fast model so it does not burn main-model tokens
- model.context_length raises the ceiling so it fires later.
https://hermes-agent.nousresearch.com/docs/developer-guide/context-compression-and-caching
Here’s the promised compiled post, will update with any new ones!
