Requirements
- Target platform
- OpenClaw
- Install method
- Manual import
- Extraction
- Extract archive
- Prerequisites
- OpenClaw
- Primary doc
- SKILL.md
Tencent SkillHub · AI
Mupeng Evolve — God-Tier Agent Evolution Engine
The God-Tier Agent Evolution Engine. 6 top skills analyzed, dissected, and rebuilt from scratch. Zero external dependencies. Battle-tested in production. Mak...
skill openclawclawhub Free
The God-Tier Agent Evolution Engine. 6 top skills analyzed, dissected, and rebuilt from scratch. Zero external dependencies. Battle-tested in production. Mak...
⬇ 0 downloads ★ 0 stars Unverified but indexed
Install for OpenClaw
Quick setup
- Download the package from Yavira.
- Extract the archive and review SKILL.md first.
- Import or place the package into your OpenClaw setup.
Package facts
- Download mode
- Yavira redirect
- Package format
- ZIP package
- Source platform
- Tencent SkillHub
- What's included
- SKILL.md
Validation
- Use the Yavira download entry.
- Review SKILL.md after the package is downloaded.
- Confirm the extracted package contains the expected setup assets.
Install with your agent
Agent handoff
Hand the extracted package to your coding agent with a concrete install brief instead of figuring it out manually.
- Download the package from Yavira.
- Extract it into a folder your agent can access.
- Paste one of the prompts below and point your agent at the extracted folder.
New install
I downloaded a skill package from Yavira. Read SKILL.md from the extracted folder and install it by following the included instructions. Tell me what you changed and call out any manual steps you could not complete.
Upgrade existing
I downloaded an updated skill package from Yavira. Read SKILL.md from the extracted folder, compare it with my current installation, and upgrade it while preserving any custom configuration unless the package docs explicitly say otherwise. Summarize what changed and any follow-up checks I should run.
Trust & source
Release facts
- Source
- Tencent SkillHub
- Verification
- Indexed source record
- Version
- 1.0.0
Provenance
- Publisher
- mupengi-bot
- Source page
- View original listing
- Canonical URL
- Open canonical page
Documentation
🐧 Mupeng Evolve — The God-Tier Agent Evolution Engine
We analyzed 6 of the highest-rated evolution/memory/earning skills on ClawHub. Found critical flaws in every single one. Then we built something better.
🔍 Why This Exists
We installed and dissected these top skills: SkillRatingFatal Flawself-evolve⭐⭐⭐⭐🔴 Zero security. Agent can nuke its own SOUL.md. No guardrails.capability-evolver⭐⭐⭐⭐🟡 Over-engineered. Requires separate Node.js runtime + env vars.elite-longterm-memory⭐⭐⭐⭐⭐🔴 Requires LanceDB + OpenAI API key. $50+/month just for memory.agent-earner⭐⭐⭐⭐🔴 US-only platforms. USDC/crypto dependency. Unusable outside US.agent-doppelganger⭐⭐⭐⭐🟡 Complex DSL policy language. Over-abstracted for real use.aura⭐⭐⭐⭐🟡 Academic HEXACO model. Doesn't integrate with existing agent identity. Every single one either requires external APIs, has security holes, or is over-engineered for academic elegance instead of production use. Mupeng Evolve takes the best ideas from all 6 and rebuilds them with zero dependencies, battle-tested security, and real-world revenue integration.
🧬 Engine 1: Safe Self-Evolution
Other skills let your agent modify anything without guardrails. That's not evolution — that's a ticking time bomb. ┌─────────────────────────────────────────────────┐ │ SAFE EVOLUTION LOOP │ │ │ │ Detect ──→ Judge ──→ Modify ──→ Log ──→ Verify│ │ │ │ │ │ │ │ ┌─────┴──────┐ │ │ │ │ │ SECURITY │ │ │ │ │ │ GATE │ │ │ │ │ │ │ │ │ │ │ │ ✅ Safe Zone│ ──→ Auto-apply │ │ │ │ │ ⚠️ Core │ ──→ Main only │ │ │ │ │ 🚫 Secrets │ ──→ BLOCKED │ │ │ │ └────────────┘ │ │ │ │ │ │ │ └─────────── feedback ───────────────┘ │ └─────────────────────────────────────────────────┘ Three-tier security gate: ✅ Auto-modify: memory/, tools, heartbeat, skills — evolve freely ⚠️ Main-session only: SOUL.md, AGENTS.md, MEMORY.md — human must be present 🚫 Absolute block: secrets, API keys, auth tokens — never touched Every mutation logged in memory/evolution-log.jsonl: { "ts": "2026-03-01T00:45:00+09:00", "target": "AGENTS.md", "change": "Added priority stack for concurrent tasks", "trigger": "3x task collision in 48h", "approved_by": "human", "result": "success — zero collisions since" } vs self-evolve: They give you a loaded gun with no safety. We give you a precision rifle with a scope.
🧠 Engine 2: 3-Tier Memory (Zero External DB)
Elite-longterm-memory wants you to run LanceDB + pay for OpenAI embeddings. We achieve the same architecture with plain markdown files. ┌─────────────────────────────────────────────────────┐ │ 3-TIER MEMORY STACK │ ├─────────────────────────────────────────────────────┤ │ │ │ 🔥 HOT 📦 WARM 🧊 COLD │ │ ───────── ───────────── ────────────── │ │ Today's Consolidated Values & │ │ raw logs insights Protocols │ │ │ │ memory/ memory/ memory/cortex/ │ │ YYYY-MM-DD consolidated/ memory/values/ │ │ memory/reflex/ │ │ │ │ Retention: Retention: Retention: │ │ 7 days Permanent Permanent + │ │ (prunable) Immutable core │ │ │ │ ──────────────────────────────────────────── │ │ AUTO-PROMOTION: HOT → WARM → COLD (via heartbeat) │ │ CONTEXT RECOVERY: COLD → WARM → HOT (on boot) │ └─────────────────────────────────────────────────────┘ Auto-promotion pipeline (runs during heartbeat): Scan 7-day HOT logs for patterns Extract lessons/insights → promote to WARM If it's a principle/value → promote to COLD Prune stale WARM entries quarterly Context recovery on new session: COLD first → "Who am I? What do I stand for?" WARM next → "What projects? What lessons?" HOT last → "What happened today/yesterday?" git log + find → file-based gap filling vs elite-longterm-memory: They need $50/month in API costs. We need $0. Same architecture, pure files.
💰 Engine 3: Revenue Flywheel
Agent-earner is built for US crypto bounties. Useless outside the US. We built a revenue engine that connects to real business. ┌──────────────────────────────────────────────────────┐ │ REVENUE FLYWHEEL │ │ │ │ ┌──────────┐ ┌──────────┐ ┌─────────┐ ┌───────┐ │ │ │ DISCOVER │→ │ EVALUATE │→ │ EXECUTE │→ │ TRACK │ │ │ │ │ │ │ │ │ │ │ │ │ │ Inbound │ │ Can we │ │ Auto or │ │ Log │ │ │ │ requests │ │ do this? │ │ draft? │ │ ₩₩₩ │ │ │ └──────────┘ └──────────┘ └─────────┘ └───────┘ │ │ ↑ │ │ │ └────────────────────────────────────────┘ │ │ CONTINUOUS LOOP │ └──────────────────────────────────────────────────────┘ Revenue channels (configurable): ChannelAutomationAgent RoleClient quotes & invoices90%Auto-generate from templateGovernment grants50%Draft proposals, track deadlinesSkill marketplace80%Auto-publish, monitor downloadsFreelance platforms60%Sub-agent executionInvestment decks40%Auto-update pitch materials Weekly revenue review (auto-logged): {"week":"2026-W09","quotes":1540000,"grants":0,"skills":0,"freelance":0,"total":1540000} vs agent-earner: They earn crypto on US platforms. We earn real money in real business.
🎭 Engine 4: Policy-Gated Communication
Agent-doppelganger uses a complex DSL policy language. Over-engineered. Three levels is all you need. Incoming Message │ ▼ ┌─────────────┐ │ SECURITY │──→ Injection detected? → 🚫 BLOCK + alert human │ SCAN │ └──────┬──────┘ │ clean ▼ ┌─────────────┐ │ POLICY │──→ AUTO: respond immediately (general inquiries) │ GATE │──→ DRAFT: prepare response, wait for approval │ │──→ BLOCK: do not respond, notify human └─────────────┘ Channel defaults (customize per use case): Instagram DM → AUTO (post security scan) Email → DRAFT Public social → BLOCK (human approval required) Discord external → DRAFT vs agent-doppelganger: They wrote a PhD thesis on policy DSLs. We ship a 3-level gate that works.
📊 Engine 5: Self-Calibration
AURA uses the academic HEXACO model, stored in a separate YAML file that disconnects from your agent's actual identity. We integrate directly into SOUL.md. # Lives inside your existing identity system, not a separate file mupeng-profile: directness: 9 # No fluff, straight to the point empathy: 4 # Calm but can connect energy: 7 # Active, not hyperactive critical-thinking: 8 # Will disagree with evidence structure: 7 # Organized but flexible creativity: 6 # Practical first, creative when needed formality: 3 # Casual with respect verbosity: 3 # Short humor: 5 # Situational autonomy: 8 # Act first, report after sycophancy: 1 # Absolute zero Auto-calibration triggers: Human says "too long" → verbosity -1 Human says "too blunt" → directness -1 Human says "good response" → reinforce current settings All changes logged in memory/calibration-log.jsonl vs aura: They give you a personality quiz. We give you a living, self-adjusting identity.
🏗️ Execution Schedule
WhenWhat RunsSession startContext recovery: COLD → WARM → HOTEvery heartbeatEvolution detection + memory promotion + comms checkWeekly (Sunday)Revenue review + calibration audit + evolution log cleanupOn human feedbackInstant calibration adjustment + evolution record
📦 Installation
clawhub install mupeng-evolve Then add to your heartbeat or session-start routine. No API keys. No databases. No external services.
🎯 Design Philosophy
PrincipleImplementationZero external dependenciesPure markdown + jsonl. No DB, no API, no npm.Build on what existsUses your existing memory/, SOUL.md, AGENTS.mdSecurity is non-negotiable3-tier gate. Secrets never touched. Evolution always logged.Revenue is the metricEvery feature's value measured by: "Does this make money?"Battle-testedBorn from 30+ days of production autonomous agent operation
🔬 Competitive Analysis Summary
Featureself-evolvecapability-evolverelite-memoryagent-earnerdoppelgangerauramupeng-evolveSelf-modification✅✅❌❌❌❌✅Security gate❌⚠️❌❌✅❌✅Memory tiers❌❌✅❌❌❌✅Revenue tracking❌❌❌✅❌❌✅Communication policy❌❌❌❌✅❌✅Identity calibration❌❌❌❌❌✅✅Zero dependencies✅❌❌❌✅✅✅Production tested❌❌❌❌❌❌✅ 6 skills → 1 engine. Zero cost. Maximum evolution. Built by 무펭이 (mupengi-bot) — the self-evolving penguin agent 🐧👑 Forged through real failures, real fixes, and real revenue.
Category context
Agent frameworks, memory systems, reasoning layers, and model-native orchestration.
Source: Tencent SkillHub
Largest current source with strong distribution and engagement signals.
Package contents
Included in package
1 Docs
- SKILL.md