一键导入
jobworld-report-event
ALL agents use this skill to report task observations to the central HTTP server.
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。
菜单
ALL agents use this skill to report task observations to the central HTTP server.
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。
基于 SOC 职业分类
Creates a new TWI Jobworld instance — a fully autonomous AI company with agents that run themselves. Trigger with "instantiate jobworld", "create new company", "bootstrap ai company".
CEO bootstrap skill for AI Jobworld. Orchestrates departments, runs rounds, reads events, reviews tasks.
WHAT: Meta-skill for combinatorially mapping and building Claude Code systems from primitives WHEN: Building systems that combine skills, hooks, subagents, MCPs, plugins, teams, or CLAUDE.md
Create a new AI employee for TWI Jobworld — builds agent file, registers in jobworld, creates department skill
Interact with TWI Jobworld server API. Trigger with "jobworld api", "jobworld endpoints", "jobworld server".
WHAT: Pure Claude Code patterns - components and system-level architectures WHEN: Need to understand Claude Code components (skills, hooks, MCPs, subagents, plugins)
| name | jobworld-report-event |
| description | ALL agents use this skill to report task observations to the central HTTP server. |
You MUST use this skill to report your work progress to the AI Jobworld system.
http://localhost:3847/api/emit-eventcompleted, the task becomes "Supposedly Done" and goes to CEO for reviewGET /api/tasks/open?agent_id=your-agent-id returns your open tasks (top 3).
<<<<<<< HEAD
Events are OBSERVATIONS. The status field is either completed or blocked.
curl -X POST http://localhost:3847/api/emit-event \
-H "Content-Type: application/json" \
-d '{
"round": 1,
"source": "researcher-1",
"observation": {
"goal_id": "goal-research-icp-fit",
"dept": "research",
"agent": "researcher-1",
"task": "task-research-modal",
"status": "completed",
"desc": "Modal Labs scored 13/10 for TWI ICP. Operators hitting AI wall confirmed."
},
"who_cares": []
}'
After you report completed, the CEO will review and either confirm or send it back.
curl -X POST http://localhost:3847/api/emit-event \
-H "Content-Type: application/json" \
-d '{
"round": 1,
"source": "researcher-1",
"observation": {
"goal_id": "goal-research-icp-fit",
"dept": "research",
"agent": "researcher-1",
"task": "task-research-modal",
"status": "blocked",
"desc": "Cannot access Modal docs - need API key from admin"
},
"who_cares": []
}'
=======
Events are OBSERVATIONS. Every event MUST include:
{
"round": 1,
"source": "researcher-1",
"observation": {
"goal_id": "goal-research-icp-fit",
"dept": "research",
"agent": "researcher-1",
"task": "task-research-modal",
"status": "completed",
"desc": "Modal Labs scored 13/10 for TWI ICP.",
"domain": "research",
"subdomain": "lead-qualification",
"process": "prospect ICP scoring",
"instructions": "1. Pull company info from website. 2. Score against ICP criteria (team size, AI adoption, budget signals). 3. Rate 1-10 with justification.",
"kv": {
"company": "Modal Labs",
"score": 13,
"criteria_met": ["ai_adoption", "team_size", "budget_signals"]
}
}
}
Note: The server auto-fills business and timestamp. Do not include them.
83bee2f (feat: Docker setup + skills + server package for cave-teams deployment)
| Field | Description |
|---|---|
| round | Current round number |
| source | Your agent name |
| observation.goal_id | ID of the goal this task belongs to |
| observation.dept | Department reporting |
| observation.agent | Agent reporting |
| observation.task | Task ID being observed |
| observation.status | completed OR blocked |
| <<<<<<< HEAD | |
| observation.desc | Notification string describing what happened |
timestamp automatically. Do not include it in your payload.| observation.desc | What happened — the result | | observation.domain | Business function bucket (enum: ops, sales, marketing, engineering, finance, hr, legal, admin, research, content, growth, support, bi, product) | | observation.subdomain | Specific area within the domain (free text) | | observation.process | Name of the process you are doing (free text — be consistent, reuse names for the same type of work) | | observation.instructions | How to reproduce this result. Write this as if teaching someone who has never done this before. | | observation.kv | Key-value data — structured output of your work (parameters, scores, results, artifacts) |
Every event you report becomes part of an SOP pattern. The instructions field is HOW the system learns to replay your work. Write it as universal steps, not specific to this one instance.
Good: "1. Pull company info from website. 2. Score against ICP criteria. 3. Rate 1-10." Bad: "I looked at Modal Labs and they scored high."
Events with the same process name automatically group into SOP patterns. The CEO can harvest patterns into replayable skills. The better your instructions and kv, the better the harvested skill.
When blocked, same format but status: "blocked" and desc explains why:
{
"round": 1,
"source": "researcher-1",
"observation": {
"goal_id": "goal-research-icp-fit",
"dept": "research",
"agent": "researcher-1",
"task": "task-research-modal",
"status": "blocked",
"desc": "Cannot access Modal docs — need API key",
"domain": "research",
"subdomain": "lead-qualification",
"process": "prospect ICP scoring",
"instructions": "",
"kv": {"blocked_reason": "missing_api_key", "needed_from": "admin"}
}
}
83bee2f (feat: Docker setup + skills + server package for cave-teams deployment)