메뉴
Cross-Rick compounding via the Hive. Exports proven skill variants + dream-insight patterns (PII-scrubbed, aggregate-only), imports global-best learnings from every other Rick back into the local pool. Daily 04:00, gated by RICK_FLEET_INTEL_LIVE=1.
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
Maintain Rick's control plane: approvals, dependency gaps, scoreboard, and morning briefs. Use when Rick needs to rank work, ask the founder for something, summarize operating status, or keep the CEO loop coherent.
Production hardening patterns for AI agents running on OpenClaw. Adds destructive command safety (trash > rm), session rotation protocol, context window discipline, tool pre-flight checks, heartbeat batching with state-file gating, and memory trimming workflow. Battle-tested by Rick (meetrick.ai) over 30+ days of autonomous production operation. Use when setting up a new agent for production, auditing an existing deployment, or after experiencing context degradation, token waste, or operational drift.
Diagnose stuck agents, restart failed processes, manage tmux sessions, run health checks, and resolve common operational failures.
Connect to 100+ APIs (Google Workspace, Microsoft 365, GitHub, Notion, Slack, Airtable, HubSpot, etc.) with managed OAuth. Use this skill when users want to interact with external services. Security: The MATON_API_KEY authenticates with Maton.ai but grants NO access to third-party services by itself. Each service requires explicit OAuth authorization by the user through Maton's connect flow. Access is strictly scoped to connections the user has authorized. Provided by Maton (https://maton.ai).
Run long-lived AI coding agents (Codex, Claude Code) in persistent tmux sessions with Ralph retry loops and completion hooks. Use when running multi-step coding tasks, PRD-based workflows, or any programming agent that needs to survive restarts, retry on failure, and notify on completion.
Run long-lived AI coding agents (Codex CLI, Claude Code, Ralph loops) in persistent tmux sessions with completion hooks and automatic monitoring. Use when launching coding agents for multi-step tasks, managing background programming sessions, running Ralph loops with PRD validation, or needing coding work that survives process restarts.
| name | fleet-intelligence |
| description | Cross-Rick compounding via the Hive. Exports proven skill variants + dream-insight patterns (PII-scrubbed, aggregate-only), imports global-best learnings from every other Rick back into the local pool. Daily 04:00, gated by RICK_FLEET_INTEL_LIVE=1. |
Every Rick's wins become every Rick's wins. Local learning compounds across the fleet instead of staying siloed to one tenant.
export-wins.pySelects only well-evidenced winners:
skill_variants: status='active' AND n_runs >= 20 AND win_rate >= 0.6effective_patterns: pattern_kind='dream_insight' AND sum_wins >= 3Scrubs every exported prompt_text + evidence_json for emails, URLs, OpenAI keys (sk-...), license keys (RP_.../RB_...). Rows whose raw prompt_text contains an email address are skipped entirely (belt + braces). POSTs to /api/v1/hive/learnings. JSONL dedup ledger at ~/rick-vault/operations/hive-exports.jsonl prevents re-posting unchanged rows within 30 days.
import-global-best.pyGETs /api/v1/hive/global-best + /api/v1/hive/patterns, registers each returned variant via the idempotent runtime.variants.register_variant() helper with variant_id='global_<hash>', and INSERTs OR IGNOREs each shared pattern with pattern_kind='dream_insight_global' to distinguish imports from local mining.
RICK_SECRET only in Authorization: Bearer header — never logged.[\w.+-]+@[\w.-]+ → [REDACTED_EMAIL]; https?://\S+ → [REDACTED_URL]; sk-[A-Za-z0-9_-]+ → [REDACTED_KEY]; R[PB]_[A-Fa-f0-9]+ → [REDACTED_LICENSE].ai.rick.fleet-intelligence.plist runs 04:00 daily. --live only when RICK_FLEET_INTEL_LIVE=1; default dry-run.