Top AI Repos — open-source AI, indexed and scored
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
Hermes-native AIOps agent for evidence-driven incident response, approval-gated remediation, and runbook learning.
| Date | Stars |
|---|---|
| 2026-07-31 | 536 |
| 2026-08-03 | 538 |
| 2026-08-06 | 538 |
Today
— stars today
This week
— stars this week
This month
— stars this month
Momentum
0.0
growth rate 0.00%/day
# RunbookHermes **基于 Hermes Agent 改造的 AIOps / SRE 事故响应 Agent：证据驱动 RCA、审批保护执行、自进化 Runbook 记忆、RAG 知识库、评测与训练数据闭环。** RunbookHermes 不是一个普通聊天机器人，也不是一个简单的监控看板。它的目标是把 Hermes Agent 的 Agent Runtime、工具系统、Memory、Skills、Gateway、多模型 Provider、上下文压缩和训练轨迹能力，改造成一个面向真实生产事故的 AIOps Agent。 它的工作方式是： ```text 告警进入 → 先查证据 → 再判断根因 → 再给行动方案 → 危险动作必须审批、checkpoint、dry-run → 执行后必须验证恢复 → 最后沉淀 memory、RAG、runbook skill、eval case、training trajectory ``` RunbookHermes 最重要的价值不是“回答一次问题”，而是让系统在每次事故后变得更懂业务： ```text 越用越懂系统 越用越懂故障模式 越用越懂服务治理规则 越用越懂团队排障习惯 ``` --- ## 1. RunbookHermes 是什么 RunbookHermes 是一个 Hermes-native 的事故响应系统。 它保留 Hermes Agent 的底层能力，并在其上增加 AIOps / SRE 领域层： ```text Hermes Agent Runtime + RunbookHermes AIOps Domain Layer = RunbookHermes ``` Hermes Agent 提供通用 Agent 能力： - Agent 主循环； - Tool Registry； - 多模型 Provider； - CLI / Gateway 入口； - MemoryProvider； - Skills； - 上下文压缩； - 会话状态； - RL 轨迹采集。 RunbookHermes 在这些能力上增加： - 事故接入； - Prometheus / Loki / Trace / Deploy evidence； - EvidenceStack 上下文； - RCA guard； - Action policy； - 审批、checkpoint、dry-run； - controlled execution； - recovery verification； - 自进化 memory； - citation RAG； - multimodal evidence； - eval benchmark； - training / RL dataset pipeline； - Web Console； - Feishu / WeCom / Alertmanager 入口。 --- ## 2. 核心原则 RunbookHermes 的设计原则非常明确。 ### Evidence first 事故处理中不能先猜根因。 RunbookHermes 必须先收集： - metrics； - logs； - traces； - deploy history； - service profile； - memory context； - RAG citation； - multimodal evidence。 然后再生成 RCA。 ### Memory is weak prior 记忆只能作为背景信息，不能替代当前证据。 ```text 历史记忆可以提醒 Agent： “这个服务以前也出现过类似故障” 但最终结论必须来自当前事故的 metrics、logs、traces、deploy evidence。 ``` ### Dangerous action must be gated 危险动作不能由模型直接执行。 包括但不限于： - rollback； - restart； - scaling mutation； - traffic switching； - config mutation； - database-affecting operation； - cache flush； - dependency failover。 RunbookHermes 的安全链路是： ```text action policy → approval → checkpoint → dry-run → controlled execution → recovery verification → audit timeline ``` ### Every incident improves the next incident 一次事故结束后，RunbookHermes 会沉淀： - incident summary； - fault pattern； - service governance rule； - team runbook habit； - generated SKILL.md； - RAG document； - eval case； - training trajectory； - reward label。 这就是 RunbookHermes 越用越懂系统的原因。 --- --- ## Product Screenshots The screenshots below show the current RunbookHermes Web Console. Put these images under `docs/assets/` and keep the file names consistent with the Markdown paths. ### AIOps Console Overview  The overview page shows the high-level AIOps control plane: incident count, pending approvals, generated skills, critical services, recommended operation flow, current capability boundaries, and a live monitoring preview. ### Realtime Monitoring System  The monitoring page provides a multi-dimensional service health view for `payment-service`, `coupon-service`, and `order-service`, including HTTP status signals, QPS, p95 latency, service topology, backend mode, and deployment state.  The lower section of the monitoring page shows log signals and trace signals. This is where RunbookHermes connects observability data to incident diagnosis instead of relying only on model guesses. ### Incident Command Center  The incident list page normalizes incidents created from Web, Alertmanager, Feishu, WeCom, or API entry points. It shows service, status, severity, root cause, creation time, and quick incident creation actions. ### Incident Detail: Evidence and Executive Summary  The incident detail page displays evidence cards from metrics, logs, and traces, plus an executive summary with root cause, recommended action, evidence IDs, confidence, and approval status. ### Incident D
Excerpt of 36,083 characters
Read on GitHub
5
Would you bet a product on this? Bounded 0–100 and slow moving.
matched fp:36490e4ddd1b2d7d, llm:description: 'Hermes-native AIOps agent for evidence-driven incident response, approval-gated remediation, and runbook learning.' Language: Python. No topics provided.
matched fp:36490e4ddd1b2d7d, llm:description: 'Hermes-native AIOps agent for evidence-driven incident response, approval-gated remediation, and runbook learning.' Language: Python. No topics provided.
matched fp:36490e4ddd1b2d7d, llm:description: 'Hermes-native AIOps agent for evidence-driven incident response, approval-gated remediation, and runbook learning.' Language: Python. No topics provided.