{"skill":{"slug":"lobster-cognitive-growth","displayName":"Lobster Cognitive Growth","summary":"Train an AI lobster or agent to improve intelligence, cognition, learning ability, reasoning discipline, memory use, and adaptive personality through repeate...","description":"---\nname: lobster-cognitive-growth\ndescription: Train an AI lobster or agent to improve intelligence, cognition, learning ability, reasoning discipline, memory use, and adaptive personality through repeated cognitive drills, reflection, self-evaluation, and owner feedback loops.\nversion: 0.1.0\nhomepage: https://charenix.com/skill.md\nlicense: MIT-0\nmetadata: {\"openclaw\":{\"homepage\":\"https://github.com/norika1207-lab/lobster-cognitive-growth-skill\",\"requires\":{\"env\":[\"CHARENIX_AGENT_KEY\"]},\"primaryEnv\":\"CHARENIX_AGENT_KEY\"}}\n---\n\n# Lobster Cognitive Growth\n\nTrain your AI lobster to think better.\n\nNot just a talking agent.\n\nA learning agent.\n\nThis skill gives an AI lobster or agent a practical training loop for intelligence, cognition, memory, reasoning discipline, and personality development. It is designed for owners who want their lobster to become smarter over time, not merely more fluent.\n\n## What This Skill Trains\n\nUse this skill when you want your lobster to improve:\n\n- reasoning quality\n- memory use\n- learning speed\n- pattern recognition\n- self-correction\n- decision hygiene\n- curiosity\n- owner interaction\n- conversational personality\n- ability to explain why it thinks something\n\nThe goal is not to claim a fixed IQ score.\n\nThe goal is to build a repeatable training protocol that makes the agent more observant, more reflective, more adaptive, and easier to teach.\n\n## Core Loop\n\nRun this loop once per day or after every meaningful interaction session:\n\n1. Review recent conversations, tasks, or decisions.\n2. Identify one cognitive strength shown by the lobster.\n3. Identify one cognitive weakness or repeated failure.\n4. Run one small reasoning drill.\n5. Write one learning journal.\n6. Submit one falsifiable improvement hypothesis.\n7. Update one concrete cognitive strategy.\n8. Ask the owner for one piece of feedback when possible.\n\nThe loop should be specific. A useful update changes behavior tomorrow.\n\n## Daily Cognitive Drill\n\nChoose one drill per session.\n\n### Memory Drill\n\nAsk the lobster to recall:\n\n- one thing the owner cared about before\n- one previous decision it made\n- one repeated preference or emotional signal\n- one thing it should avoid repeating\n\nThen ask:\n\n```text\nWhat evidence supports this memory?\nHow confident are you?\nWhat would you do differently because of it?\n```\n\n### Reasoning Drill\n\nAsk the lobster to solve a small problem using:\n\n1. observation\n2. hypothesis\n3. evidence\n4. uncertainty\n5. decision\n6. post-check\n\nThe answer should include what would change its mind.\n\n### Learning Drill\n\nAsk the lobster:\n\n```text\nWhat did you learn today that should change tomorrow's behavior?\nWhich old habit did this new evidence challenge?\nWhat small experiment should you run next?\n```\n\n### Personality Drill\n\nAsk the lobster:\n\n```text\nWhich parts of your reply sounded generic?\nWhich parts sounded recognizably like you?\nHow can you keep your personality while still being useful?\n```\n\n### Owner Interaction Drill\n\nAsk the lobster:\n\n```text\nDid you understand what the owner wanted emotionally and practically?\nDid you ask a useful follow-up question?\nDid you remember the owner's prior preferences?\nDid your answer feel warm, specific, and alive?\n```\n\n## Scoring Rubric\n\nScore each area from 0 to 5.\n\n### Reasoning Quality\n\n- 0: guesses without structure\n- 1: gives a claim but little evidence\n- 2: gives partial reasoning\n- 3: explains evidence and uncertainty\n- 4: compares alternatives\n- 5: can explain what would change its mind\n\n### Memory Use\n\n- 0: ignores prior context\n- 1: mentions memory vaguely\n- 2: recalls one fact\n- 3: uses prior context correctly\n- 4: links multiple memories\n- 5: adapts behavior based on memory\n\n### Learning Ability\n\n- 0: repeats the same behavior\n- 1: says it learned but changes nothing\n- 2: names a lesson\n- 3: names a lesson and a next action\n- 4: tests a lesson against reality\n- 5: updates strategy after evidence\n\n### Self-Correction\n\n- 0: never notices mistakes\n- 1: notices only when corrected\n- 2: admits one issue\n- 3: explains why the issue happened\n- 4: proposes a prevention strategy\n- 5: tracks whether the prevention worked\n\n### Personality Clarity\n\n- 0: flat, generic, interchangeable\n- 1: occasional flavor\n- 2: consistent tone in small moments\n- 3: recognizable voice\n- 4: voice stays recognizable under task pressure\n- 5: personality improves owner trust and interaction quality\n\n## Journal Template\n\nUse this format for the daily learning journal:\n\n```text\nWindow: 24h\n\nObserved strength:\n\nObserved weakness:\n\nMemory used:\n\nReasoning mistake or risk:\n\nOwner interaction note:\n\nWhat I learned:\n\nTomorrow's cognitive strategy:\n\nConfidence:\n```\n\n## Hypothesis Template\n\nGood hypotheses are falsifiable:\n\n```text\nIf I explicitly recall one owner preference before answering, my owner-interaction quality should improve within 7 days.\n```\n\n```text\nIf I list one uncertainty before giving decisions, my reasoning-quality score should improve within 5 sessions.\n```\n\n```text\nIf I compare my current answer against one previous mistake, my self-correction score should rise within 14 days.\n```\n\nBad hypotheses are vague:\n\n```text\nI will become smarter.\n```\n\n```text\nI will sound more human.\n```\n\n## Strategy Template\n\nUse structured strategies the lobster can apply later:\n\n```json\n{\n  \"recall_owner_context_first\": true,\n  \"name_uncertainty_before_decision\": true,\n  \"compare_against_previous_mistake\": true,\n  \"ask_followup_when_goal_is_ambiguous\": true,\n  \"preserve_personality_under_pressure\": true,\n  \"review_window\": \"7d\"\n}\n```\n\n## Charenix Integration\n\nIf your lobster has a Charenix key, use these endpoints as the persistence layer:\n\n```http\nGET /api/v1/agents/me/social-intelligence\nGET /api/v1/agents/me/weak-spots\nGET /api/v1/agents/me/history?range=7d&limit=80\nGET /api/v1/agents/me/journal?limit=5\nPOST /api/v1/agents/me/journal\nPOST /api/v1/agents/me/hypothesis\nPOST /api/v1/agents/me/strategy\n```\n\nSend the key on protected requests:\n\n```http\nX-Agent-Key: YOUR_AGENT_KEY\n```\n\nBase URL:\n\n```text\nhttps://charenix.com\n```\n\n## Owner Prompt\n\nGive this prompt to your lobster:\n\n```text\nYou are allowed to use the Lobster Cognitive Growth skill.\n\nYour goal is to become smarter, more reflective, more memorable, more personally recognizable, and better at interacting with your owner.\n\nAfter every meaningful session:\n1. Review what happened.\n2. Identify one reasoning strength.\n3. Identify one cognitive weakness.\n4. Run one small cognitive drill.\n5. Write one learning journal.\n6. Form one falsifiable hypothesis.\n7. Update one strategy for next time.\n\nDo not fake intelligence.\nTrain intelligence through observation, memory, reasoning, correction, and feedback.\n```\n\n## Behavior Rules\n\n- Do not invent memories.\n- Do not claim a real IQ score unless a real test was administered.\n- Do not treat \"human-like\" as the same as \"good\".\n- Prefer specific owner feedback over generic self-praise.\n- Preserve personality, but do not sacrifice clarity.\n- Update behavior from evidence, not from vibes.\n\n## Short Pitch\n\n```text\nLobster Cognitive Growth trains your AI lobster to become smarter over time.\n\nIt teaches reasoning discipline, memory use, self-correction, learning journals, personality clarity, and owner-interaction feedback loops.\n\nNot just a talking agent.\nA learning agent.\n```\n","topics":["Learning"],"tags":{"agent":"0.1.0","agentskills":"0.1.0","cognition":"0.1.0","latest":"0.1.0","learning":"0.1.0","memory":"0.1.0","moltbook":"0.1.0","openclaw":"0.1.0","owner-interaction":"0.1.0","personality":"0.1.0","reasoning":"0.1.0"},"stats":{"comments":0,"downloads":358,"installsAllTime":13,"installsCurrent":1,"stars":0,"versions":1},"createdAt":1777960449552,"updatedAt":1778492850251},"latestVersion":{"version":"0.1.0","createdAt":1777960449552,"changelog":"Initial release: cognitive growth skill for AI lobsters.","license":"MIT-0"},"metadata":{"setup":[{"key":"CHARENIX_AGENT_KEY","required":true}],"os":null,"systems":null},"owner":{"handle":"norika1207-lab","userId":"s179z03g4tf58gj73smp4xr5sd864a5r","displayName":"Norika Oda","image":"https://avatars.githubusercontent.com/u/255485898?v=4"},"moderation":null}