Tommy-ywCommunity·132 Agent Skills Included

RunbookHermes

AI incident response agent that diagnoses outages and learns from every fix

Automates SRE incident response by collecting metrics, logs, traces, and deploy history before proposing root causes. Gates dangerous actions like rollbacks and restarts behind approvals, checkpoints, dry-runs, and recovery verification. Turns every resolved incident into reusable runbooks, memory, RAG knowledge, and training data so the system improves over time.

539|39

↗

npx skills add Tommy-yw/RunbookHermes --all -g -y

All Skills in This Repository (132)

Pure Emerald Level Indicators

📦 In Repo

Tommy-yw

parallel-cli

Execute asynchronous web searches and monitor research tasks via Parallel CLI, outputting structured JSON results.

gitnexus-explorer

Index codebases into a knowledge graph and serve an interactive web UI.

qmd

Search local knowledge bases with keyword, semantic, and LLM reranking.

drug-discovery

Search ChEMBL for bioactive compounds and calculate drug-likeness.

duckduckgo-search

Runs DuckDuckGo text, news, image, and video searches via CLI or Python.

scrapling

Extract web data with stealthy multi-method fetching for Python.

bioinformatics

Fetch domain-specific bioinformatics reference material for computational biology tasks.

domain-intel

Perform passive domain reconnaissance using Python stdlib.

docker-management

Automate Docker container operations, image management, and Dockerfile optimization.

inference-sh-cli

Run AI applications via the infsh CLI without a GPU.

sherlock

Query 400+ social networks to locate accounts linked to a username.

oss-forensics

Investigate open-source supply chain attacks across seven phases using GitHub and BigQuery.

Community

Advanced

Frequently Asked Questions

FAQPage Schema

How to install RunbookHermes?▼

Run `npx skills add Tommy-yw/RunbookHermes --all -g -y` in your terminal to install the full skill suite globally.

What does RunbookHermes do for incident response?▼

It ingests alerts, gathers evidence from Prometheus, Loki, traces, and deploy history, then proposes a root cause and action plan. Risky fixes require human approval, checkpoints, and dry-runs before execution.

How does RunbookHermes prevent unsafe automated actions?▼

Every dangerous operation goes through an action policy, human approval, checkpoint creation, dry-run, controlled execution, and post-action recovery verification, all recorded in an audit timeline.

Does RunbookHermes get better over time?▼

Yes. Each incident produces a summary, fault pattern, generated runbook skill, RAG document, eval case, and training trajectory, so the agent learns your systems and team habits.

Can RunbookHermes connect to Feishu, WeCom, or Alertmanager?▼

Yes. It accepts incidents from the web console, Alertmanager, Feishu, WeCom, and API entry points, and includes a settings page showing integration readiness.