Exécutez n'importe quel Skill dans Manus
en un clic

Exécutez n'importe quel Skill dans Manus en un clic

Commencer

capy-cortex

Étoiles132

Forks25

Mis à jour20 mars 2026 à 11:09

Autonomous learning system - learns from mistakes, reflects on sessions, and gets smarter over time. The AI brain.

Installation

Installer avec Codex ou Claude Copiez ce prompt, collez-le dans Codex, Claude ou un autre assistant, puis laissez-le vérifier la page du skill et l'installer pour vous.

Exécuter dans Manus

Source

happycapy-ai

happycapy-ai/Happycapy-skills

Ouvrir le dépôt GitHub Voir les dépôts du créateur

Téléchargement

Exécuter dans Manus

Métiers associésSOC

Basé sur la classification professionnelle SOC

Développeurs de logicielsProfessions informatiques et mathématiques·SOC 15-1252

Explorateur de fichiers

67 fichiers

SKILL.md

readonly

name	capy-cortex
description	Autonomous learning system - learns from mistakes, reflects on sessions, and gets smarter over time. The AI brain.

Capy Cortex - Autonomous Learning System

You have a persistent learning brain powered by SQLite + FTS5 + sklearn TF-IDF. Knowledge is automatically loaded via hooks. This file describes manual operations.

Architecture

Database: ~/.claude/skills/capy-cortex/cortex.db (SQLite + FTS5 + WAL)
Hooks (automatic, never call manually):
- SessionStart: Loads anti-patterns, preferences, principles
- UserPromptSubmit: Retrieves task-relevant rules via FTS5
- PreToolUse(Bash): Blocks known dangerous commands
- PostToolUseFailure: Records errors as anti-patterns
- Stop: Extracts corrections and preferences from conversation
Scripts (for manual/scheduled use):
- cortex.py: Core engine (retrieve, add rules, stats)
- reflect.py: Deep session analysis
- consolidate.py: Cluster rules into principles (sklearn)
- bootstrap.py: Mine historical sessions

Manual Commands

# Check system health
python3 ~/.claude/skills/capy-cortex/scripts/cortex.py stats

# Retrieve rules for a topic
python3 ~/.claude/skills/capy-cortex/scripts/cortex.py retrieve "react typescript"

# Add a rule manually
python3 ~/.claude/skills/capy-cortex/scripts/cortex.py add-rule "Always use TypeScript strict mode" "best_practice"

# Add an anti-pattern
python3 ~/.claude/skills/capy-cortex/scripts/cortex.py add-ap "Never force push to main" "critical"

# Add a preference
python3 ~/.claude/skills/capy-cortex/scripts/cortex.py add-pref "User prefers functional components over class components"

# Run consolidation (clusters rules into principles)
python3 ~/.claude/skills/capy-cortex/scripts/consolidate.py

# Retrain TF-IDF model
python3 ~/.claude/skills/capy-cortex/scripts/cortex.py retrain

# Apply confidence decay
python3 ~/.claude/skills/capy-cortex/scripts/cortex.py decay

How It Learns

Automatic (via hooks): Errors are captured, corrections noted, preferences extracted
Reflection: Deep analysis of session transcripts extracts patterns
Consolidation: sklearn clustering groups similar rules into principles
Decay: Old, unreinforced rules fade; validated rules strengthen
Retrieval: Two-stage FTS5 + TF-IDF returns only relevant knowledge (O(1) context)

Plus depuis ce dépôt

même dépôt

apricot-ritual-pptx

happycapy-ai/Happycapy-skills

Reusable fixed-style PPTX generation skill for the Apricot Ritual Index visual brand — warm apricot background (#FDDEB6), cocoa-brown typography (#6C4832), refined minimalist editorial poster aesthetic, elegant serif titles, narrow-sans labels, centered word stacks, thin horizontal rules, sparse image placement, widescreen 16:9 editable .pptx output. Use this skill whenever the user mentions Apricot Ritual Index, asks for a warm apricot minimalist presentation, wants a refined product-poster or ritual-index style deck, or requests a presentation using the Apricot Ritual Index visual system. Also trigger when the user says 'apricot ritual', 'ritual index deck', 'apricot minimalist slides', 'product poster presentation', 'ritual index pptx', 'editorial poster deck', 'typographic poster presentation', or 'warm apricot slides' in any context.

2026-07-10132

azure-host-pptx

happycapy-ai/Happycapy-skills

Reusable fixed-style PPTX generation skill for the Azure Host Forum visual brand — sky-blue background (#65AEFF), deep green typography (#07461E), olive side-panel accent (#728E03), cream portrait/detail color (#F1F0EA), bold contemporary industry sharing event aesthetic, oversized condensed grotesk titles, strong left-text / right-speaker-column composition, circular portrait frames, widescreen 16:9 editable .pptx output. Use this skill whenever the user mentions Azure Host Forum, asks for a presentation or deck in the Azure Host Forum style, wants a forum or industry-sharing deck, speaker event deck, expert talk slides, panel presentation, creative industry event deck, founder briefing slides, or any event/forum-style PowerPoint. Also trigger when the user says 'azure host forum', 'forum deck', 'speaker event slides', 'industry sharing presentation', 'host forum pptx', 'panel deck', 'workshop deck', or 'creative industry event slides' in any context.

2026-07-10132

botanical-blur-pptx

happycapy-ai/Happycapy-skills

Reusable fixed-style PPTX generation skill for the Botanical Blur System visual brand — warm cream (#F7F4EA) editorial layouts, botanical motion-blur textures, sky blue (#3D8FD9), deep green (#0E5A3D), muted coral (#D9573F), dark slate (#2B302F), white leaf-outline motifs, varied slide family with full-bleed texture slides, curved bands, rounded panels, and structured grids. Widescreen 16:9 editable .pptx output. Use this skill whenever the user mentions Botanical Blur System, asks for a botanical motion-blur or organic editorial deck, wants cream editorial layouts with botanical textures, or asks for a botanical-style workshop, strategy, brand, product, or storytelling deck. Also trigger for 'botanical blur', 'botanical blur system', 'organic editorial deck', 'botanical pptx', 'cream botanical slides', 'blurred botanical presentation', or 'premium organic presentation'.

2026-07-10132

coral-geyser-pptx

happycapy-ai/Happycapy-skills

Reusable fixed-style PPTX generation skill for the Coral Geyser Notice visual brand — coral-red background (#E95140), warm stone typography (#E7D8D1), bold modern visitor-notice and design-forward travel guide aesthetic, heavy geometric sans titles, strong split-screen compositions, widescreen 16:9 editable .pptx output. Use this skill whenever the user mentions Coral Geyser Notice, asks for a presentation or deck in the Coral Geyser style, wants a visitor notice deck, travel guide deck, field guide presentation, park-hours-poster-style PowerPoint, route card deck, or mentions coral-red with warm stone typography for slides. Also trigger when the user says 'coral geyser', 'notice deck', 'visitor guide deck', 'field guide slides', 'travel notice presentation', 'park guide deck', or 'route card presentation' in any context.

2026-07-10132

olive-rose-pptx

happycapy-ai/Happycapy-skills

Reusable fixed-style PPTX generation skill for the Olive Rose Salon visual brand — dark olive-brown background (#372E19), pale pink typography (#FFE1FC), high-fashion art exhibition poster aesthetic, large serif title type, widescreen 16:9 editable .pptx output. Use this skill whenever the user mentions Olive Rose Salon, asks for a presentation or deck in the Olive Rose style, wants an exhibition poster deck, needs a high-fashion poster-led PowerPoint, or mentions the olive-brown/pale-pink color scheme for slides. Also trigger when the user says 'olive rose', 'salon deck', 'poster presentation', or 'fashion deck' in any context.

2026-07-10132

signal-green-pptx

happycapy-ai/Happycapy-skills

Reusable fixed-style PPTX generation skill for the Signal Green Atlas visual brand — fluorescent green (#0CDA76) dominant background, black (#000000) typography and rules, black-and-white imagery only, bold avant-garde editorial atlas aesthetic, oversized uppercase grotesk headlines, strict two-column information grids, sharp rectangular photo windows, black circular bullets, deliberate asymmetry, widescreen 16:9 editable .pptx output. Use this skill whenever the user mentions Signal Green Atlas, asks for a fluorescent-green editorial presentation, wants an avant-garde atlas or guidebook-style deck, requests a bold black-and-green presentation, or needs a deck using strict information grids and monochrome imagery. Also trigger for 'signal green', 'signal green atlas', 'fluorescent green deck', 'atlas pptx', 'black and green presentation', 'avant-garde atlas deck', 'guidebook style slides', or 'editorial atlas presentation'.

2026-07-10132

Capy Cortex - Autonomous Learning System

You have a persistent learning brain powered by SQLite + FTS5 + sklearn TF-IDF. Knowledge is automatically loaded via hooks. This file describes manual operations.

Architecture

Database: ~/.claude/skills/capy-cortex/cortex.db (SQLite + FTS5 + WAL)

Hooks (automatic, never call manually):

SessionStart: Loads anti-patterns, preferences, principles
UserPromptSubmit: Retrieves task-relevant rules via FTS5
PreToolUse(Bash): Blocks known dangerous commands
PostToolUseFailure: Records errors as anti-patterns
Stop: Extracts corrections and preferences from conversation

Scripts (for manual/scheduled use):

cortex.py: Core engine (retrieve, add rules, stats)
reflect.py: Deep session analysis
consolidate.py: Cluster rules into principles (sklearn)
bootstrap.py: Mine historical sessions

Manual Commands

# Check system health python3 ~/.claude/skills/capy-cortex/scripts/cortex.py stats # Retrieve rules for a topic python3 ~/.claude/skills/capy-cortex/scripts/cortex.py retrieve "react typescript" # Add a rule manually python3 ~/.claude/skills/capy-cortex/scripts/cortex.py add-rule "Always use TypeScript strict mode" "best_practice" # Add an anti-pattern python3 ~/.claude/skills/capy-cortex/scripts/cortex.py add-ap "Never force push to main" "critical" # Add a preference python3 ~/.claude/skills/capy-cortex/scripts/cortex.py add-pref "User prefers functional components over class components" # Run consolidation (clusters rules into principles) python3 ~/.claude/skills/capy-cortex/scripts/consolidate.py # Retrain TF-IDF model python3 ~/.claude/skills/capy-cortex/scripts/cortex.py retrain # Apply confidence decay python3 ~/.claude/skills/capy-cortex/scripts/cortex.py decay

How It Learns

Automatic (via hooks): Errors are captured, corrections noted, preferences extracted

Reflection: Deep analysis of session transcripts extracts patterns

Consolidation: sklearn clustering groups similar rules into principles

Decay: Old, unreinforced rules fade; validated rules strengthen

Retrieval: Two-stage FTS5 + TF-IDF returns only relevant knowledge (O(1) context)