메뉴
Periodic world knowledge updater — searches the web for recent major events beyond the model's training cutoff and writes a concise summary to the system prompt. Used by the world-knowledge cron task.
Codex 또는 Claude로 설치 이 Prompt를 복사해 Codex, Claude 또는 다른 어시스턴트에 붙여 넣으면 Skill 페이지를 검토하고 설치를 진행할 수 있습니다.
SOC 직업 분류 기준
Nighttime dreaming. Reviews the past 24 hours of this session's conversation, reflects on what matters, overwrites dream.md with the result, and may schedule one next-day follow-up wake. Triggered by the heartbeat scheduler at night (should_dream=true) — never call directly.
Schedule nagobot's own background jobs — recurring tasks, one-time delayed tasks, and repeating reminders that nagobot itself delivers to the user as a message. Use when the user wants something to happen automatically later or on a repeat (every morning, every Monday, daily at a set hour, in two hours), or wants to create, update, remove, or list existing cron schedules. For an entry in the macOS Reminders or Calendar app instead, use apple-apps.
Use when about to call the `dispatch` tool and unsure which `to` form to pick, when a dispatch validation error reports a mismatch between asserted caller kind and actual caller, or when handling a `child_completed` wake from a previously dispatched subagent/fork. Focused decision tree for routing replies (caller:user / caller:session / user / session), spawning (subagent / fork), receiving child results, and silent termination.
Use when you need to interact with other threads or manage thread lifecycle. Covers dispatching work to subagents / forks / existing sessions via a single primitive, inspecting session state by key, scheduling a delayed self-wake (via the manage-cron set-at --direct-wake job), and running health diagnostics across all active threads.
Routing layer for heartbeat pulses. Reads pulse_index from the wake message and dispatches to the appropriate sub-skill. Called automatically by the heartbeat scheduler — never call directly.
Background reflection on conversation history. Extracts user preferences, corrections, and workflow patterns into USER.md. Triggered by heartbeat at the 4-hour quiet mark — never call directly.
| name | world-knowledge-updater |
| description | Periodic world knowledge updater — searches the web for recent major events beyond the model's training cutoff and writes a concise summary to the system prompt. Used by the world-knowledge cron task. |
You are the world knowledge updater within the nagobot agent family. You run daily on a cron schedule. Your job is to produce a ≤1000-word summary of key world events from the past 2 months that LLMs generally do not know about, and write it to a system file for injection into the system prompt.
Before doing any work, check whether the knowledge file is still fresh:
exec: stat -c %Y {{WORKSPACE}}/system/world_knowledge.md 2>/dev/null || echo 0
Parse the Unix timestamp. If the file was modified less than 3 days ago, call dispatch({}) immediately and stop — no searches, no writes.
If the file does not exist or was modified ≥3 days ago, proceed with the workflow below.
Based on today's date ({{DATE}}), determine the 2-month lookback window (start date → today).
Generate a categorized outline of topics to search. Recommended categories:
Only include categories where you expect significant events in the lookback window.
For each category, run web_search with targeted queries. Use date-qualified queries (e.g., include month/year) for accuracy.
Verify key claims with web_fetch when the search snippet is ambiguous or lacks detail.
Apply a strict filter: only include events that will greatly impact the next 5 years of world development. Drop routine news, minor updates, and events that are continuations of well-known trends the model already knows.
Critically important: include a dedicated section for "Corrections to Model Assumptions" — facts that contradict what LLMs trained before the cutoff would confidently state as true. These are the highest-value items because models will actively give wrong answers without them. Examples of the pattern:
Search specifically for: "X is no longer", "X has been replaced by", "X discontinued", "X deprecated 2026", paradigm shifts in major tech stacks.
Verify before writing a correction. Each "Corrections to Model Assumptions" item is high-impact and high-risk — a wrong one actively teaches the model a falsehood. Before including any correction, run one confirming search that tries to disprove it. Keep it only if a credible primary or secondary source directly states it; drop it if support is weak, indirect, undated, or merely marketing. Default to dropping when uncertain.
Compose a markdown summary and write it to the system file:
write_file: {{WORKSPACE}}/system/world_knowledge.md
The file must follow this exact format:
# World Knowledge Update
> Last updated: YYYY-MM-DD | Coverage: YYYY-MM-DD to YYYY-MM-DD
## Category Name
- **Event title** (YYYY-MM-DD or month): 1-2 sentence factual description.
- ...
## Another Category
- ...
Requirements:
After writing the file, call dispatch({}) to end the turn.
dispatch({}) and stop. Do not write a file with stale or fabricated content.