AI SRE Incident Response
Guides agents through incidents in LLM services: outages, quality drops, safety issues and cost spikes.
- Repository
- sickn33/agentic-awesome-skills
- Stars
- 49,023
- Licence
- MIT
- Files
- 1 file
What it does
This skill gives the agent a structured playbook for handling production incidents in AI systems, such as a model going down, answers getting worse, safety filters firing too often, or token spending spiking. Once installed, the agent can suggest severity levels, alert rules for Prometheus, and step-by-step runbooks that a responder can follow, including fallback routing and rollbacks. It is documentation-style guidance; the file itself contains no executable scripts.
Great for
- On-call platform or SRE engineers building monitoring and response for LLM-powered services
- Teams that run self-hosted or gateway-routed models on Kubernetes and need a starting runbook
- Engineers who want example Prometheus alert rules for AI latency, cost and guardrail metrics
Skip it if
- Anyone without authorized access to a production environment, since the runbooks issue kubectl commands that can change live deployments
- Teams expecting a verified quality measurement: the skill states that its groundedness proxy and thresholds are illustrative and must be calibrated before use
Try asking
- “Write Prometheus alert rules for token cost spikes per tenant in our LLM gateway”
- “Draft a runbook for when our model provider goes down and traffic must fail over”
- “Create a severity matrix for AI quality and safety incidents for our on-call team”
- “Review our hallucination-rate alert and suggest how to tune its threshold”
Install in 3 steps
- 1
Download the skill
The ZIP contains the 1 original files, the MIT licence and an attribution note.
ai-sre-incident-response.zip - 2
Add it to your agent
Claude Code (all projects)
mkdir -p ~/.claude/skills && unzip ai-sre-incident-response.zip -d ~/.claude/skills/Claude Code / Agent SDK (this project only)
mkdir -p .claude/skills && unzip ai-sre-incident-response.zip -d .claude/skills/Claude apps (web/desktop): Settings → Capabilities → Skills → Upload skill, and choose the ZIP.
- 3
Ask for the task
No special command needed — the agent reads the skill's description and loads it when your request matches. Try one of the prompts above.
Get better results
- Treat every threshold and command as a template: calibrate against your own SLOs and metric units before applying anything to production.
- Ask the agent to confirm deployment-freeze controls actually enforce the freeze, since the skill warns an annotation alone does not.
- Keep the generated runbooks in version control and have a human review any kubectl or rollback command before running it.
What's inside
- SKILL.md
Redistributed unmodified from sickn33/agentic-awesome-skills under the MIT licence. The assessment was drafted with AI from the skill's files on 2026-10-29. Skills can instruct your agent to run code — read SKILL.md and any scripts before installing.