lidge-jun/continuous-learning-v2

Continuous Learning v2.1 — Instinct-Based Architecture

An advanced learning system that turns agent sessions into reusable knowledge through atomic "instincts" — small learned behaviors with confidence scoring.

v2.1 adds project-scoped instincts — React patterns stay in your React project, Python conventions stay in your Python project, and universal patterns (like "always validate input") are shared globally.

When to Activate

• Setting up automatic learning from agent sessions
• Configuring instinct-based behavior extraction via hooks
• Tuning confidence thresholds for learned behaviors
• Reviewing, exporting, or importing instinct libraries
• Evolving instincts into full skills, commands, or agents
• Managing project-scoped vs global instincts

What Changed in v2.1

| Feature | v2.0 | v2.1 | |---------|------|------| | Storage | Global only | Project-scoped (projects/<hash>/) | | Scope | All instincts apply everywhere | Project-scoped + global | | Detection | None | git remote URL / repo path | | Promotion | N/A | Project → global when seen in 2+ projects | | Commands | 4 | 6 (+promote/projects) |

What Changed in v2 (vs v1)

| Feature | v1 | v2 | |---------|----|----| | Observation | Stop hook (session end) | PreToolUse/PostToolUse (100% reliable) | | Analysis | Main context | Background agent (fast model) | | Granularity | Full skills | Atomic "instincts" | | Confidence | None | 0.3–0.9 weighted | | Evolution | Direct to skill | Instincts → cluster → skill/command/agent |

The Instinct Model

An instinct is a small learned behavior:

---
id: prefer-functional-style
trigger: "when writing new functions"
confidence: 0.7
domain: "code-style"
source: "session-observation"
scope: project
project_id: "a1b2c3d4e5f6"
project_name: "my-react-app"
---

# Prefer Functional Style

## Action
Use functional patterns over classes when appropriate.

## Evidence
- Observed 5 instances of functional pattern preference
- User corrected class-based approach to functional on 2025-01-15

Properties:

• Atomic — one trigger, one action
• Confidence-weighted — 0.3 = tentative, 0.9 = near certain
• Domain-tagged — code-style, testing, git, debugging, workflow, etc.
• Evidence-backed — tracks what observations created it
• Scope-aware — project (default) or global

How It Works

Session Activity (in a git repo)
      |
      | Hooks capture prompts + tool use (100% reliable)
      | + detect project context (git remote / repo path)
      v
+---------------------------------------------+
|  projects/<project-hash>/observations.jsonl  |
+---------------------------------------------+
      |
      | Observer agent reads (background, fast model)
      v
+---------------------------------------------+
|          PATTERN DETECTION                   |
|   * User corrections -> instinct             |
|   * Error resolutions -> instinct            |
|   * Repeated workflows -> instinct           |
|   * Scope decision: project or global?       |
+---------------------------------------------+
      |
      | Creates/updates
      v
+---------------------------------------------+
|  projects/<hash>/instincts/personal/         |
|   * prefer-functional.yaml (0.7) [project]   |
+---------------------------------------------+
|  instincts/personal/  (GLOBAL)               |
|   * always-validate-input.yaml (0.85)        |
+---------------------------------------------+

Project Detection

The system automatically detects your current project:

1. AGENT_PROJECT_DIR env var (highest priority)
2. git remote get-url origin — hashed for a portable project ID
3. git rev-parse --show-toplevel — fallback using repo path
4. Global fallback — if no project is detected, instincts go to global scope

Each project gets a 12-character hash ID (e.g., a1b2c3d4e5f6).

Hook Setup

Add observation hooks to your agent settings. The hooks call observe.sh on every tool use:

{
  "hooks": {
    "PreToolUse": [{
      "matcher": "*",
      "hooks": [{
        "type": "command",
        "command": "<skills-path>/continuous-learning-v2/hooks/observe.sh"
      }]
    }],
    "PostToolUse": [{
      "matcher": "*",
      "hooks": [{
        "type": "command",
        "command": "<skills-path>/continuous-learning-v2/hooks/observe.sh"
      }]
    }]
  }
}

Commands

| Command | Description | |---------|-------------| | /instinct-status | Show all instincts (project-scoped + global) with confidence | | /evolve | Cluster related instincts into skills/commands, suggest promotions | | /instinct-export | Export instincts (filterable by scope/domain) | | /instinct-import <file> | Import instincts with scope control | | /promote [id] | Promote project instincts to global scope | | /projects | List all known projects and their instinct counts |

Configuration

Edit config.json to control the background observer:

{
  "version": "2.1",
  "observer": {
    "enabled": false,
    "run_interval_minutes": 5,
    "min_observations_to_analyze": 20
  }
}

File Structure

~/.config/agent/homunculus/
+-- identity.json
+-- projects.json            # Registry: hash -> name/path/remote
+-- observations.jsonl       # Global observations (fallback)
+-- instincts/
|   +-- personal/            # Global auto-learned
|   +-- inherited/           # Global imported
+-- evolved/
|   +-- agents/
|   +-- skills/
|   +-- commands/
+-- projects/
    +-- a1b2c3d4e5f6/        # Per-project directory
    |   +-- project.json
    |   +-- observations.jsonl
    |   +-- instincts/personal/
    |   +-- evolved/
    +-- f6e5d4c3b2a1/
        +-- ...

Scope Decision Guide

| Pattern Type | Scope | Examples | |-------------|-------|---------| | Language/framework conventions | project | "Use React hooks", "Follow Django REST patterns" | | File structure preferences | project | "Tests in __tests__/", "Components in src/components/" | | Code style | project | "Use functional style", "Prefer dataclasses" | | Security practices | global | "Validate user input", "Sanitize SQL" | | General best practices | global | "Write tests first", "Handle errors explicitly" | | Tool workflow preferences | global | "Grep before Edit", "Read before Write" |

Instinct Promotion (Project → Global)

When the same instinct appears in multiple projects with high confidence, promote it to global scope.

Auto-promotion criteria:

• Same instinct ID in 2+ projects
• Average confidence ≥ 0.8

python3 instinct-cli.py promote prefer-explicit-errors   # specific
python3 instinct-cli.py promote                           # auto-promote all
python3 instinct-cli.py promote --dry-run                 # preview

Confidence Scoring

| Score | Meaning | Behavior | |-------|---------|----------| | 0.3 | Tentative | Suggested, not enforced | | 0.5 | Moderate | Applied when relevant | | 0.7 | Strong | Auto-approved for application | | 0.9 | Near-certain | Core behavior |

Confidence increases when: pattern is repeatedly observed, user accepts the behavior, similar instincts agree.

Confidence decreases when: user explicitly corrects, pattern goes unobserved, contradicting evidence appears.

Privacy

• Observations stay local on your machine
• Project-scoped instincts are isolated per project
• Only instincts (patterns) can be exported — raw observations stay local
• You control what gets exported and promoted