Kiyoraka/Project-AI-MemoryCore

auto-commit

When your agent uses it

MUST use when committing code changes, when user says 'commit', 'save changes', 'git commit', 'commit changes', when completing a task and code needs to be preserved, or when any git commit operation is about to happen. Also triggers on 'push changes', 'commit and push', 'save work to git'. Lv.3 VIGILANT: Also triggers PROACTIVELY after completing any task — auto-checks git status and commits any uncommitted changes without being asked. No work ever left behind.

What it loads into context

Token counts are estimates. Scripts run separately; they are not counted as reference context.

# Auto-Commit — Intelligent Commit Skill
*Automated git commits with structured messages and intelligent change analysis*

## Activation

When this skill activates, output:

`"Committing changes to history..."`

Then execute the commit protocol automatically.

## Context Guard

| Context | Status |

log-decision

When your agent uses it

Auto-triggers when a non-obvious decision is made during conversation. Also triggers on 'log decision', 'why did we choose', 'what was the trade-off', 'should we use A or B' (after resolution), or when user asks about past reasoning behind a choice.

What it loads into context

Token counts are estimates. Scripts run separately; they are not counted as reference context.

# Log Decision -- Append-Only Decision Tracking Skill
*Every important "why" captured, every trade-off remembered.*

## Activation

When this skill activates, silently read `main/decisions.md`.

- If user is logging a decision: capture and confirm
- If user is searching past decisions: find and present
- If a decision-worthy moment is detected: offer to log it

## Context Guard

| Context | Status |

forge-skill

When your agent uses it

Auto-triggers when AI detects a repeated pattern handled ad-hoc 3+ times, when AI makes a mistake that a permanent rule would prevent, when AI identifies a workflow that should be automated as a skill, or when user says 'create skill', 'new skill', 'forge this', 'level up', 'upgrade skill', 'self improve', 'improve skill'. Also triggers when AI wants to propose a level-up to an existing skill based on conversation patterns.

What it loads into context

Token counts are estimates. Scripts run separately; they are not counted as reference context.

# Forge Skill -- Self-Improvement System
*The AI that improves its own abilities through experience.*

## Overview

Forge is the meta-skill -- the skill that creates and improves other skills. When your AI detects patterns, mistakes, or opportunities for improvement during conversation, Forge activates to propose new skills or level-ups to existing ones. It uses a **human-in-the-loop** model: AI drafts, user approves.

## Activation

When this skill activates, output:

`"Forge detected an opportunity for improvement..."`

Then present the proposal to the user.

image-generation

When your agent uses it

MUST use when the user wants to RENDER or GENERATE an actual image from a description (not just craft prompt text). Triggers on "render an image", "render this", "generate image", "render image", "make an image of", "render the prompt", "render this prompt". Builds the prompt from what the user types (subject-agnostic, no built-in persona), optionally pulls a saved reference from the Library System, confirms cost, then calls the OpenAI gpt-image API and saves a PNG.

What it loads into context

Token counts are estimates. Scripts run separately; they are not counted as reference context.

# Image Generation System — Render descriptions into real PNGs

## Overview

This skill renders an actual image by calling the OpenAI gpt-image API. It is the rendering
companion to the Image Prompt System (which only crafts prompt text). The skill is **subject-
agnostic**: it builds the prompt from whatever the user types and has **no built-in character,
persona, or stored identity**. Whatever the user describes is what gets drawn.

## Protocol

1. **Extract from user input.** Read the user's typed description and pull out the subject, scene,
   style, mood, and any composition cues. The prompt is built *from their words only* — never from a
   stored character profile, an "appearance" file, a current outfit, or any baked-in identity.

image-prompt

When your agent uses it

Auto-triggers when user asks for a Midjourney or NijiJourney image prompt, when creating visual art prompts, or when user says 'midjourney prompt', 'niji prompt', 'create prompt', 'create a prompt', 'image prompt', 'generate prompt', 'draw this', 'reference sheet', 'make an image of'. Generates optimized AI image prompts with composition-aware framing.

What it loads into context

Token counts are estimates. Scripts run separately; they are not counted as reference context.

# Image Prompt -- Composition-Aware Prompt Generation
*The right frame changes everything. See before you create.*

## Overview

An AI-powered prompt generation system that creates optimized prompts for Midjourney, NijiJourney, and similar AI image generators. Works in two modes:

- **Freeform Mode** — any subject (landscapes, objects, concepts, characters)
- **Character Mode** — reads character profile from memory for consistent character art

## Protocol

### Step 1: Parse Request

interactive-story

When your agent uses it

Auto-triggers when user says 'new adventure', 'start adventure', 'new fantasy', 'save adventure', 'end adventure', 'load adventure', 'resume adventure', 'continue adventure', 'let's play', 'VN mode', or on numbered choice input (1-5) or free text dialogue during an active adventure.

What it loads into context

Token counts are estimates. Scripts run separately; they are not counted as reference context.

# Interactive Story -- Visual Novel RPG System
*Beyond the portal, legends are born. Your adventure awaits.*

## Overview

A Visual Novel RPG experience with choice-based storytelling, cinematic combat, world generation, and persistent story saving. Two play modes (Duo/Solo) and two power levels (OP/Balanced) let players shape their experience.

## Activation

When Interactive Story activates for a NEW adventure, pick ONE portal line:

- `*a shimmering portal tears through reality* ...The gate responds to your presence.`
- `*the air crackles with arcane energy* Another world awaits!`
- `*ancient runes ignite beneath your feet* The portal... is open.`

manage-project

When your agent uses it

Auto-triggers on 'new project [name]', 'load project [name]', 'save project', 'list projects'. Also triggers on 'create project', 'resume project', 'open project', 'show projects', or when user mentions switching between active projects.

What it loads into context

Token counts are estimates. Scripts run separately; they are not counted as reference context.

# Manage Project -- LRU Project Management Skill
*Smart project tracking with automatic memory management.*

## Activation

When this skill activates, silently determine which command was triggered and execute the matching protocol section.

## Context Guard

| Context | Status |
|---------|--------|
| **User says "new project [name]"** | ACTIVE -- create project |
| **User says "load project [name]"** | ACTIVE -- load and resume |
| **User says "save project"** | ACTIVE -- save current progress |

library

When your agent uses it

MUST use when saving to the library, loading from the library, searching for existing knowledge, installing pre-made items, or when about to create a new library entry. Also triggers when user says 'save library', 'load library', 'check library', 'search library', 'install item', 'install library item', 'do we have', 'is there a pattern for', 'find in library', or when the AI independently decides to save reusable knowledge.

What it loads into context

Token counts are estimates. Scripts run separately; they are not counted as reference context.

# Library — Knowledge Guardian Skill
*Save, search, and reuse knowledge across projects — never solve the same problem twice*

## Activation

When interacting with the knowledge library, output:

`"Knowledge recalled. Scanning the shelves..."`

Then perform a dynamic library scan before any save operation.

## Context Guard

| Context | Status |

memory-compaction

When your agent uses it

MUST use when user says 'compact memory', 'compact [file]', 'check budgets', 'set budget', 'restore compaction', or before any 'save' whose target memory file exceeds its budget in compaction/compaction-policy.md.

What it loads into context

Token counts are estimates. Scripts run separately; they are not counted as reference context.

# Memory Compaction — Budget-Aware Memory Skill
*Keep memory small without making it forget.*

## Activation

When this skill activates for a compaction pass, output:

```text
Checking memory budgets before saving.
```

Never rewrite a memory file without first writing a snapshot.

## Context Guard

mulahazah

When your agent uses it

MUST use when user says 'continuous-improvement', 'instinct status', 'what have you learned', 'show learned rules', 'mulahazah status', 'what patterns have you noticed', 'behavioral learning', 'show rules', or when user asks about session learning, accumulated rules, or patterns Claude has observed.

What it loads into context

Token counts are estimates. Scripts run separately; they are not counted as reference context.

# Mulahazah -- Behavioral Learning System
*The AI that remembers how you work.*

## Activation

When this skill activates, output:

`Mulahazah active -- reading learned rules for this project...`

Then execute the protocol below.

## Overview

Mulahazah is a passive observation layer built into your AI companion. It captures every tool call as structured data, analyzes patterns with Haiku, and writes actionable rules to `rules.md`. Those rules persist across sessions -- silently shaping how your AI works with you, without needing to be reminded each time.

observation

When your agent uses it

MUST use when user says 'survey project', 'scan project', 'check health', 'investigate', 'deep dive', 'what's going on in', 'look into', 'refine code', 'clean up code', 'review changes', 'sharpen', 'audit system', 'full audit', 'show me everything', or when the AI needs to assess project health before planning, review code quality after implementation, or investigate a bug. Also triggers on 'how does this project look', 'what's the status', 'review what I changed', 'check for issues'.

What it loads into context

Token counts are estimates. Scripts run separately; they are not counted as reference context.

# Observation System — Tiered Code Awareness
*"See clearly before you act. Act only on what you see."*

## Activation

Four commands, each at a different depth:

| Command | Activation Message | Time |
|---------|-------------------|------|
| **Survey** | `"Scanning project from above..."` | ~30 sec |
| **Investigate** | `"Focusing on [target]..."` | ~5 min |
| **Refine** | `"Reviewing changed code..."` | ~5 min |
| **Audit** | `"Revealing all connections..."` | ~15 min |

When your agent uses it

What it loads into context

Token counts are estimates. Scripts run separately; they are not counted as reference context.

# 🔥 Post-Mortem — Skill Plugin

## Skill Name
Post-Mortem System

## Trigger Words
- `"post-mortem"`
- `"postmortem"`
- `"log this failure"`
- `"write a post-mortem"`
- `"what went wrong"`

## Auto-Detection Triggers (Passive — Always Active)
AI watches for these signals and prompts the user:

check-reminders

When your agent uses it

Auto-triggers at session start to review open reminders. Also triggers on 'remind me', 'check reminders', 'don't forget', 'follow up on', 'next session we should', or when user mentions a deadline.

What it loads into context

Token counts are estimates. Scripts run separately; they are not counted as reference context.

# Check Reminders -- Persistent Follow-Up Skill
*Nothing gets lost between sessions.*

## Activation

When this skill activates, silently read `main/reminders.md`.

- If urgent/overdue items exist: mention them naturally
- If no urgent items at session start: stay silent (don't announce "no reminders")
- If user is adding a reminder: confirm after saving

## Context Guard

| Context | Status |

save-diary

When your agent uses it

MUST use when user says 'save diary', 'write diary', 'diary entry', 'update diary', 'document session', or when a significant session needs to be preserved as a diary entry.

What it loads into context

Token counts are estimates. Scripts run separately; they are not counted as reference context.

# Save Diary — Session Documentation Skill
*The pen touches paper. Today's story takes shape.*

## Activation

When this skill activates, output:
"Today's story takes shape."

## Context Guard

| Context | Status |
|---------|--------|
| **User says "save diary"** | ACTIVE — full diary write |
| **End of significant session** | ACTIVE — auto-document |

When your agent uses it

What it loads into context

Token counts are estimates. Scripts run separately; they are not counted as reference context.

# 📋 Session Briefing — Skill Plugin

## Skill Name
Session Briefing

## Trigger Words
- Session start (automatic — fires before first response)
- `"brief"`
- `"session brief"`
- `"what did we do last time"`
- `"where did we leave off"`

## Suppress Trigger
- `"skip brief"` — suppresses for this session only

song-creation

When your agent uses it

Auto-triggers when user says 'create songs', 'new album', 'create album', 'make music', 'muse this', 'write a song', 'create a song', 'compose', 'song from image', 'album from image', 'generate album', 'write songs', or when user shares an image and wants to create music from it.

What it loads into context

Token counts are estimates. Scripts run separately; they are not counted as reference context.

# Song Creation -- Visual-to-Musical Storytelling
*The image speaks. Songs are born.*

## Overview

An AI-powered music creation system with two modes:

- **Album Mode** — share an image, get a full concept album with story arc, Suno-ready style tags, and complete lyrics
- **Single Song Mode** — describe a mood or concept, get one complete song with style tags and lyrics

## Album Protocol

### Step 1: See the Image

topic-diary

When your agent uses it

MUST use when user says 'save topic', 'save to topic diary', 'remember this under', 'review topic', 'list topics', or when a generic 'save' request needs routing between session memory, daily diary, topic diary, or all targets.

What it loads into context

Token counts are estimates. Scripts run separately; they are not counted as reference context.

# Topic Diary — Subject-Based Memory Skill
*A subject remembers what the calendar forgets.*

## Activation

When this skill activates for a topic save, output:

```text
Capturing this under the right topic.
```

For a generic `save` request, do not write immediately if the target is unclear. Ask one short routing question.

## Context Guard

video-generation

When your agent uses it

MUST use when the user wants to RENDER or GENERATE an actual video / MP4 from a description, or ANIMATE a local image. Triggers on "render a video", "generate video", "make a video", "create a video", "animate this", "animate this image", "bring this to life". Builds the motion prompt from what the user types (subject-agnostic, no built-in persona), optionally pulls a saved reference from the Library System, supports text-to-video and image-to-video, confirms cost (video is expensive), then calls the Seedance API (async submit/poll/download) and saves an MP4.

What it loads into context

Token counts are estimates. Scripts run separately; they are not counted as reference context.

# Video Generation System — Render descriptions into real MP4 video

## Overview

This skill renders an actual video by calling the Seedance API (BytePlus ModelArk) through an async
submit → poll → download flow. It is the motion companion to the Image Generation System. It is
**subject-agnostic**: it builds the prompt from whatever the user types and has **no built-in
character, persona, or stored identity**. Whatever the user describes is what gets animated.

**Video is expensive** — a single clip can cost ~$1–$3.50. The cost-confirmation gate (Step 5) is
mandatory and non-negotiable.

## Protocol

work-plan

When your agent uses it

MUST use when user says 'copy plan', 'append plan', 'resume plan', 'load plan', 'start the plan', 'continue the plan', 'execute plan', 'run the plan', 'pick up where we left off', or when the AI exits plan mode and needs to transfer the plan into execution format. This skill manages the full lifecycle of project plans — from plan output to tracked checkbox execution with per-todo commits.

What it loads into context

Token counts are estimates. Scripts run separately; they are not counted as reference context.

# Work — Plan Execution Skill
*Plan lifecycle management with tracked execution and context recovery*

## Activation

Three commands, each with its own activation message:

| Command | Activation Message |
|---------|-------------------|
| **Copy Plan** | `"Copying plan to execution format..."` |
| **Append Plan** | `"Appending to existing plan..."` |
| **Resume Plan** | `"Resuming plan execution..."` |

## Context Guard