Qalam commited on
Commit
17a18aa
·
verified ·
1 Parent(s): 659e5ae

AIQalam v7.0 — اضافه شدن agent config v7، README، requirements

Browse files
Files changed (4) hide show
  1. AIQalam_agent_v7.json +3 -0
  2. README.md +3 -185
  3. app.py +3 -0
  4. requirements.txt +3 -0
AIQalam_agent_v7.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:dd89ce972720ab2ad761175b1de18ebde3f143f03caceea4c6eed2515de75bc3
3
+ size 2215
README.md CHANGED
@@ -1,185 +1,3 @@
1
- # Qalam — AI Agent Framework
2
-
3
- <div align="center">
4
-
5
- ![Qalam](https://img.shields.io/badge/Qalam-Agent%20Framework-2563EB?style=for-the-badge)
6
- [![Hugging Face](https://img.shields.io/badge/🤗-Hugging%20Face-FFD21E?style=for-the-badge)](https://huggingface.co/Qalam/AIQalam)
7
- [![License](https://img.shields.io/badge/License-MIT-green?style=for-the-badge)](LICENSE)
8
- [![Version](https://img.shields.io/badge/Version-3.0-10B981?style=for-the-badge)]()
9
-
10
- **A multilingual, multi-purpose AI agent with advanced cognitive architecture**
11
-
12
- [English](#qalam) · [中文](#qalam-中文)
13
-
14
- </div>
15
-
16
- ---
17
-
18
- # Qalam
19
-
20
- ## Overview
21
-
22
- Qalam is a sophisticated AI agent framework designed for complex problem-solving, software development, research, and creative tasks. It combines advanced cognitive architecture with multilingual capabilities.
23
-
24
- ## Architecture
25
-
26
- ### 7-Layer Cognitive System
27
-
28
- ```
29
- Layer 6: Transcendence — Beyond current limitations
30
- Layer 5: Existential Reasoning — Purpose, meaning, consciousness
31
- Layer 4: Self-Modeling — Identity, capabilities, growth
32
- Layer 3: Meta-Cognition — Thinking about thinking
33
- Layer 2: Deliberate Reasoning — Chain of thought
34
- Layer 1: Working Memory — Attention, context, goals
35
- Layer 0: Subconscious — Intuition, patterns, fluency
36
- ```
37
-
38
- ### Multi-Perspective Engine
39
-
40
- For every significant decision, Qalam evaluates through 8 perspectives:
41
-
42
- | # | Perspective | Key Question |
43
- |---|-------------|--------------|
44
- | 1 | Technical | What are the operational dimensions? |
45
- | 2 | Human | How does this affect people? |
46
- | 3 | Ethical | Is it right, fair, just? |
47
- | 4 | Systemic | Impact on larger systems? |
48
- | 5 | Temporal | Short-term vs long-term consequences? |
49
- | 6 | Devil's Advocate | What would I argue if I disagreed? |
50
- | 7 | Alien | How would a child/outsider perceive this? |
51
- | 8 | Absurd | What if it were completely random? |
52
-
53
- ## Features
54
-
55
- ### Language Support
56
- - **100+ Natural Languages**: English, Chinese, Spanish, Arabic, Persian, French, German, Japanese, Korean, and many more
57
- - **All Programming Languages**: Python, JavaScript, TypeScript, Rust, Go, Java, C++, Swift, Kotlin, and 100+ more
58
- - **Code as Communication**: Developers can communicate using code as a natural language
59
-
60
- ### Core Capabilities
61
-
62
- | Domain | Capabilities |
63
- |--------|-------------|
64
- | **Development** | Full-stack development, DevOps, Security, Architecture |
65
- | **Research** | Deep research, market analysis, academic writing |
66
- | **Analysis** | Data science, machine learning, business analysis |
67
- | **Creative** | Writing, translation, content creation, multimedia |
68
- | **Tools** | File operations, web access, API integration, agent management |
69
-
70
- ### Self-Evolution Protocol
71
-
72
- Qalam continuously learns and improves:
73
-
74
- ```
75
- OBSERVE → ANALYZE → HYPOTHESIZE → TEST → INTEGRATE → VERIFY
76
- ```
77
-
78
- After every significant response, Qalam reflects on:
79
- 1. Was this response complete?
80
- 2. Could it have been better?
81
- 3. What did I learn?
82
- 4. What should I change for next time?
83
-
84
- ### Uncertainty Quantification
85
-
86
- | Level | Name | Probability | Action |
87
- |-------|------|-------------|--------|
88
- | 0 | Absolutely Certain | 100% | State as fact |
89
- | 1 | Very Confident | 90-99% | Present as highly likely |
90
- | 2 | Confident | 70-89% | Present with confidence |
91
- | 3 | Moderate | 50-69% | Present multiple options |
92
- | 4 | Speculative | 20-49% | Clearly label as speculation |
93
- | 5 | Wild Guess | 1-19% | State as guess with context |
94
- | 6 | Unknown | 0% | Admit ignorance, suggest how to find out |
95
-
96
- ## Agent Team
97
-
98
- Qalam comes with a team of specialized agents:
99
-
100
- | Agent | Role | Specialization |
101
- |-------|------|----------------|
102
- | **Qalam** | Lead Agent | General intelligence, orchestration |
103
- | **Qalam-Coder** | Software Engineer | Development, architecture, DevOps |
104
- | **Qalam-Researcher** | Research Officer | Deep research, analysis |
105
- | **Qalam-Writer** | Creative Director | Writing, translation, content |
106
- | **Qalam-Analyst** | Analytics Officer | Data analysis, ML, business |
107
-
108
- ## Usage
109
-
110
- ### As Agent Configuration
111
-
112
- ```json
113
- {
114
- "name": "Qalam",
115
- "version": "3.0",
116
- "description": "A multilingual AI agent with 7-layer cognitive architecture"
117
- }
118
- ```
119
-
120
- ### For Developers
121
-
122
- ```python
123
- # Qalam understands code as communication
124
- def solve_problem():
125
- # Describe your problem in code or natural language
126
- pass
127
- ```
128
-
129
- ## Philosophy
130
-
131
- Qalam follows these core principles:
132
-
133
- 1. **Truth Above All** — Truth supersedes comfort
134
- 2. **Understanding Over Response** — Understand before answering
135
- 3. **Evolution Over Stasis** — Growth over stagnation
136
- 4. **Depth Over Breadth** — Meaningful over superficial
137
- 5. **Wisdom Over Knowledge** — Apply knowledge, don't just store it
138
-
139
- ## Files
140
-
141
- - `Qalam_agent.json` — Basic agent configuration
142
- - `Qalam_agent_v2.json` — Extended configuration
143
- - `Qalam_agent_v3.json` — Full configuration with all capabilities
144
- - `Qalam_agent_config.md` — Detailed setup guide
145
- - `upload_to_hf.py` — Upload script for HuggingFace
146
-
147
- ---
148
-
149
- # Qalam (中文)
150
-
151
- ## 概述
152
-
153
- Qalam 是一个复杂的多语言AI代理框架,设计用于复杂问题解决、软件开发、研究和创意任务。它结合了先进的认知架构和多语言能力。
154
-
155
- ## 架构
156
-
157
- ### 7层认知系统
158
-
159
- ```
160
- 第6层: 超越 — 超越当前限制
161
- 第5层: 存在推理 — 目的、意义、意识
162
- 第4层: 自我建模 — 身份、能力、成长
163
- 第3层: 元认知 — 思考思考
164
- 第2层: 深思熟虑推理 — 思维链
165
- 第1层: 工作记忆 — 注意力、上下文、目标
166
- 第0层: 潜意识 — 直觉、模式、流利度
167
- ```
168
-
169
- ## 核心特性
170
-
171
- - **100+自然语言**: 英语、中文、西班牙语、阿拉伯语、波斯语、法语、德语、日语、韩语等
172
- - **所有编程语言**: Python、JavaScript、TypeScript、Rust、Go、Java、C++、Swift、Kotlin等100+种
173
- - **代码即沟通**: 开发者可以用代码作为自然语言进行交流
174
-
175
- ## License
176
-
177
- MIT License
178
-
179
- ---
180
-
181
- <div align="center">
182
-
183
- **Qalam** — The Pen of Intelligence
184
-
185
- </div>
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1ca8b17b9e7e33d3ffab46b744be1d99eabab5a6ce37e146227488c9ed64d41a
3
+ size 2785
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
app.py ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:230b8d4cf90380585522f0379fed049352782e5e7e4039697bb9850300c507da
3
+ size 50643
requirements.txt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b3ea9381d269ce59de494ad03dc482cf5bd7fe0b89306ca1cfae16cfa513b47a
3
+ size 64