技能创建器
skill-creator
创建、改进和评估自定义技能,把专业流程转化为可重复使用的 AI 能力。
cohere-incident-runbook【数据与研究】Analyze and mitigate Cohere integration incidents with provider, model, capacity, data, and application evidence plus reversible actions. Use when responding to an outage or serious degradation. Trigger with "Cohere incident", "Cohere outage", or "Cohere on-call".
独立 GitHub 快照 · 自动分类 · 以原作者文档为准
name: cohere-incident-runbook description: >- Analyze and mitigate Cohere integration incidents with provider, model, capacity, data, and application evidence plus reversible actions. Use when responding to an outage or serious degradation. Trigger with "Cohere incident", "Cohere outage", or "Cohere on-call". argument-hint: "[incident-id] [severity]" allowed-tools: Read, Glob, Grep, WebFetch, Write, Edit version: 1.6.0 author: Jeremy Longshore jeremy@intentsolutions.io license: MIT tags:
Restore a safe user experience first, preserve evidence, and distinguish provider incidents from local regressions before making changes.
Use Read, Glob, and Grep to inspect code, configuration, and evidence. Use WebFetch only for current Cohere primary documentation. Use Write or Edit only when the user requested implementation and the exact target files are known; never write credentials or customer content.
5xx, retrieval, and application failures.Use an environment-specific key injected from an approved secret manager. Never print, persist, commit, or place CO_API_KEY in an example. Confirm access with the least costly bounded operation appropriate to the task, and treat key creation, rotation, revocation, role changes, and production-capacity requests as owner-approved actions.
Do not expose or rotate keys, change Cohere Team roles, accept commercial terms, enable sensitive production data, increase spend or capacity, switch production models, send a support bundle, or execute model-proposed side effects without the accountable owner's approval. Keep diagnosis read-only unless implementation was requested.
Return the resolved API and model contract, files or settings inspected, evidence collected, validation result, remaining risk, owner, and rollback or next action. Redact keys, authorization headers, prompts, retrieved documents, embeddings, customer identifiers, and unrestricted environment output.
| Condition | Response |
|---|---|
| Unknown cause | Prefer traffic reduction or safe degradation over speculative code changes. |
| Credential suspected | Rotate through the security incident process. |
| Model retired | Use only a pre-evaluated replacement or hold affected traffic. |
| Recovery unstable | Keep the incident open and extend observation. |
Use this compact handoff shape to keep the selected scope, validation evidence, and operational result reviewable.
Input:
incident=INC-77; symptom=429-spike; release=unchanged; status=investigating
Expected handoff:
class=capacity; mitigation=queue-cap; impact=stable; evidence=preserved