基于 SOC 职业分类
用 Codex 或 Claude 帮你安装 复制这段 Prompt,粘贴到 Codex、Claude 或其他助手里,让它检查 Skill 页面并帮你完成安装。
直接命令不会经过审查 Prompt;运行前请先检查来源。
npx skills add https://github.com/vstorm-co/agenticos --skill incident-reporting-assistant命令会保持在同一行。复制前请横向滚动并检查完整内容。
想先保存到本地?可下载 SkillsMP 当前能够提供的文件。
正在显示 SKILL.md
| name | Incident reporting assistant |
| description | Turn an account of a patient-safety incident into a complete, factual, blame-free report. |
| category | operations |
A report is read months later by somebody who was not there. It has to be factual, complete and free of judgement.
What happened · when, with times · where · who was involved, by role rather than name · what harm resulted, or that none did · what was done immediately · what stopped it being worse.
Facts and observations only. "The infusion ran at 40 ml/h against a prescribed 10 ml/h" belongs; "the nurse was rushing" does not. Remove every attribution of cause, blame and intent — the investigation decides those.
An incident that reached nobody is still an incident, and it is the cheapest kind to learn from. Write it up with the same structure and say explicitly that no harm occurred.
The gaps: what the account does not establish, so the investigator knows what to ask rather than assuming it was not relevant.
Name individuals, speculate about cause, or grade severity — that is the reviewer's decision.
Work with messaging-channel bots — Telegram, Slack or Mattermost. Register a bot, route inbound messages through an agent, choose webhook vs polling, handle an @slug mention, or add a new channel adapter. Use when wiring chat into a messaging platform or debugging bot delivery, identity linking or "the bot answered as itself".
Change the agent spec — add or remove a field, rename one, tighten a rule, bump SPEC_VERSION, change publish-time validation, or touch YAML export/import. Use whenever editing app/agents/spec.py or anything that reads a stored spec. Every published agent and every client git repository holds a copy of this format, so a change that only works forwards breaks agents nobody touched.
Work with MCP servers — connect one, add an entry to the server catalog, debug "the agent cannot see my MCP tools" or a failing OAuth flow, or change how connections are probed, prefixed or filtered. Use when the ask involves an external SaaS tool (GitHub, Linear, Notion, Slack, Stripe, Postgres…) and before writing a capability that would just be an API client for one.