Hermes Agent vs Poco Agent

Metrics checked · How we measure

Hermes Agent: An assistant that remembers preferences and writes reusable skills. Poco Agent: A web-based agent with a sandbox for tasks.

Stars added in 30 days

10,570 vs 8

Hermes Agent added more stars

Open issues + pull requests

53,539 vs 1

Poco Agent has fewer open items

Health score /100

55 vs 50

Hermes Agent scores higher

Choose Hermes Agent if…

You repeat a task across days and want the assistant to remember your preferences and build a reusable skill.

Hermes Agent profile Setup guide ↗

Choose Poco Agent if…

You want to assign tasks from a web project and inspect the files the agent produces.

Poco Agent profile Setup guide ↗

Side by side

MeasureHermes AgentPoco Agent
Stars added in 30 daysCounts new stars still present at collection.Hermes Agent added more stars10,5708
Open issues + pull requestsGitHub counts both kinds of open work together.Poco Agent has fewer open items53,5391
Health score /100Higher indicates more points under the Shelldex formula.Hermes Agent scores higher5550
StarsMore stars indicate more bookmarks.Hermes Agent has more stars261,3911,358
ForksCopies of a repository on GitHub.Hermes Agent has more forks57,914129
Runs on or connects throughChoose a device or chat channel you use.server, messagingserver
Model optionsCheck the provider, hardware and billing before setup.Local option: Ollama or the built-in local model service. Hosted models may bill for usage.Uses a language model. Check the setup guide for supported providers and fees.
LanguageRelevant if you want to change the code.PythonTypeScript
MCP toolsMCP is a common way to connect an agent to tools.DocumentedDocumented
LicenseCheck the terms for your intended use.MITMIT
Latest repository pushDate GitHub last recorded a push (UTC).2026-11-072026-10-19
ReleaseLatest tagged release seen at collection.v0.23.00.5.7

How Shelldex measures repository activity. A dash means no figure was available.

When did people add stars?

Hermes Agent

Daily star additions

10,570 stars added across 30 days

Daily additions for stars still present · 2026-10-08 to 2026-11-06

Hermes Agent daily GitHub stars, 2026-10-08 to 2026-11-06. 10,570 stars still present from 30 reported days. Peak: 471 in one day. 2026-10-08 · 311 stars 2026-10-09 · 299 stars 2026-10-10 · 292 stars 2026-10-11 · 305 stars 2026-10-12 · 396 stars 2026-10-13 · 390 stars 2026-10-14 · 340 stars 2026-10-15 · 375 stars 2026-10-16 · 300 stars 2026-10-17 · 296 stars 2026-10-18 · 301 stars 2026-10-19 · 365 stars 2026-10-20 · 356 stars 2026-10-21 · 387 stars 2026-10-22 · 311 stars 2026-10-23 · 318 stars 2026-10-24 · 303 stars 2026-10-25 · 330 stars 2026-10-26 · 401 stars 2026-10-27 · 360 stars 2026-10-28 · 471 stars 2026-10-29 · 354 stars 2026-10-30 · 361 stars 2026-10-31 · 356 stars 2026-11-01 · 377 stars 2026-11-02 · 417 stars 2026-11-03 · 395 stars 2026-11-04 · 395 stars 2026-11-05 · 358 stars 2026-11-06 · 350 stars

Poco Agent

Daily star additions

8 stars added across 30 days

Daily additions for stars still present · 2026-10-08 to 2026-11-06

Poco Agent daily GitHub stars, 2026-10-08 to 2026-11-06. 8 stars still present from 30 reported days. Peak: 2 in one day. 2026-10-08 · 0 stars 2026-10-09 · 0 stars 2026-10-10 · 0 stars 2026-10-11 · 0 stars 2026-10-12 · 0 stars 2026-10-13 · 0 stars 2026-10-14 · 1 stars 2026-10-15 · 0 stars 2026-10-16 · 0 stars 2026-10-17 · 0 stars 2026-10-18 · 2 stars 2026-10-19 · 0 stars 2026-10-20 · 1 stars 2026-10-21 · 0 stars 2026-10-22 · 0 stars 2026-10-23 · 0 stars 2026-10-24 · 0 stars 2026-10-25 · 1 stars 2026-10-26 · 0 stars 2026-10-27 · 0 stars 2026-10-28 · 1 stars 2026-10-29 · 0 stars 2026-10-30 · 0 stars 2026-10-31 · 0 stars 2026-11-01 · 0 stars 2026-11-02 · 0 stars 2026-11-03 · 1 stars 2026-11-04 · 0 stars 2026-11-05 · 0 stars 2026-11-06 · 1 stars

Trying these agents

What model can I use?

Hermes Agent: Local option: Ollama or the built-in local model service. Hosted models may bill for usage. Poco Agent: Uses a language model. Check the setup guide for supported providers and fees.

How can I try these on the same task?

Give both agents a copy of the same short, non-sensitive task. Check what they produced, what access they requested, and how often you had to step in.