Hermes Agent Model Card
这是可索引项目报告证据页：它保留 Hermes Agent Model Card 的源材料入口、机制线索和限制提醒；正文仍需 reader/editor 与 academic public-copy review 后才能当作最终结论引用。
Hermes Agent Model Card
One Sentence
Hermes Agent remains the strongest personal-agent runtime signal in the corpus: memory, skills, messaging, automations, and trajectory generation in one public system.
Three Sentences
It is not just a demo agent; the public repository still exposes persistent knowledge, autonomous skill creation, session search, user modeling, messaging gateways, and scheduled automations. That makes it a central answer to the user’s product question: which systems are usable agents rather than only papers or lists? The 2026-06-17 morning authenticated packet shows the runtime keeps moving at public GitHub scale, so the metadata surface has to stay fresh.
Model Card
| Field | Value |
|---|---|
| Repository | NousResearch/hermes-agent |
| Source | raw-github/nousresearch_hermes-agent.md |
| Category | Self-improving personal agent runtime |
| Pattern | persistent knowledge -> skill creation/improvement -> session search -> user modeling -> messaging and automation runtime |
| Evidence | Authenticated GitHub API snapshot, 2026-06-17 08:29 +0800 |
Teaching Use
Use Hermes Agent to teach the runtime surface of self-evolving agents. Improvement loops need memory, skill reuse, user modeling, search over prior sessions, and channels where the agent can actually act.
Evidence And Limits
The raw capture now reflects a GitHub metadata packet observed on 2026-07-05: 209,194 stars, 38,173 forks, 14,427 commits, 8,600 open issues, and 17,193 open pull requests. This packet is fresher than the previous authenticated packet at 2026-07-05 01:38 +0800 where a delta was observed. This run did not execute the repository locally, validate workflows end to end, or independently rerun benchmark claims. Product, memory, benchmark, and automation claims therefore remain repository-scoped unless separately tested.