A local-first AI CLI toolkit for developers. Shell commands, PR reviews, SQL generation, and image generation in one terminal. Bring your own API key. Pay once. Use forever.
| Package | Install | What it does |
|---|---|---|
| AI Shell Hub | pip install ai-shell-hub |
Natural language to shell commands, error diagnosis |
| AI PR Review | pip install ai-pr-review |
Automated code review with security pattern scanning |
| AI SQL | pip install ai-sqlx |
Natural language to SQL (MySQL, PostgreSQL, etc.) |
| AI Image CLI | pip install ai-img-cli |
Image generation via DALL-E 3 from terminal |
Each tool has a free tier. The Bundle unlocks all four Pro tiers.
# Install individual tools
pip install ai-shell-hub
pip install ai-pr-review
pip install ai-sqlx
pip install ai-img-cli
# Configure your API key
export OPENAI_API_KEY="sk-..."
ai --configai "show disk usage"
ai -e "docker-compose up -d"
ai -f "docker: command not found"
ai -Hpr-review https://github.com/example/repo/pull/123
pr-review https://github.com/example/repo/pull/123 --deepai-sqlx "find users who registered in the last 7 days"
ai-sqlx -d postgres "total orders per customer"ai-img "a cat in space wearing a hoodie"
ai-img -s 1792x1024 "wide landscape"| Tier | Price | Scope |
|---|---|---|
| Bundle (recommended) | $10 lifetime | All 4 tools |
| Shell Hub Pro | $7 lifetime | Shell AI only |
| PR Review Pro | $7 lifetime | Code review only |
| SQL Pro | $7 lifetime | SQL generation only |
| Image CLI Pro | $4 lifetime | Image generation only |
| Free | $0 | 1 trial across all tools |
Network: ERC20 (Ethereum)
Asset: USDT
Address: 0xafc32581a9e4ea30aa03cb8ef5879c2366d35f46
Important:
- Send only USDT on the ERC20 network. Other networks may result in loss.
- Transactions are irreversible. Verify the address before sending.
- This project does not custody user funds.
- Payment purchases a local software license, not a service subscription.
After sending USDT, run the claim command with your transaction hash:
# Bundle (all 4 tools)
aitools claim <tx_hash>
# Individual tools
ai claim <tx_hash>
pr-review --claim <tx_hash>
ai-sqlx claim <tx_hash>
ai-img claim <tx_hash>The CLI queries the Etherscan API to verify the transaction, then generates a local activation key. No remote server is involved in activation.
You get 3 free calls per day across all tools. After that:
You have used your free runs for today.
Upgrade to AI Developer Tools Bundle:
- AI Shell
- AI PR Review
- AI SQL
- AI Image CLI
Price: $10 lifetime
Payment: USDT ERC20 / Alipay / WeChat
Run: aitools price
Already paid?
Run: aitools claim <tx_hash>
flowchart LR
A[Developer] --> B[pip install]
B --> C[Local CLI]
C --> D[AI Shell]
C --> E[PR Review]
C --> F[SQL Generator]
C --> G[Image CLI]
C --> H[Your API Key]
C --> I[Local License]
J[USDT Payment] --> K[Etherscan Verify]
K --> I
sequenceDiagram
participant U as User
participant C as CLI
participant W as Wallet
participant E as Etherscan
participant L as Local License
U->>C: aitools price
C->>U: Show Bundle price + USDT address
U->>W: Send USDT ERC20
W->>U: Transaction hash
U->>C: aitools claim
C->>E: Verify transaction
E->>C: Confirm amount + receiver
C->>L: Activate Pro locally
C->>U: Pro activated, all tools unlocked
flowchart TD
A[Collect Issues] --> B[Classify Severity]
B --> C{Priority}
C -->|P0/P1| D[Auto Fix]
C -->|P2/P3| E[Review PR]
D --> F[Run Tests]
F --> G{Pass?}
G -->|Yes| H[Release Patch]
G -->|No| I[Rollback]
H --> J[Changelog]
J --> K[Daily Report]
I --> K
E --> K
flowchart TD
A[Trend Scan] --> B[Draft Content]
B --> C[GitHub / Landing Page]
C --> D[pip install]
D --> E[Free Trial 3x/day]
E --> F[Upgrade Prompt]
F --> G[USDT Payment]
G --> H[Auto Activation]
H --> I[Bundle Upsell]
I --> J[Feedback]
J --> K[Maintenance]
K --> C
- All tools run locally. No telemetry, analytics, or usage data is collected.
- LLM queries are sent to OpenAI (or your configured provider). Your API key is stored locally and never transmitted to this project.
- Etherscan API calls query public blockchain data only.
- No user code is uploaded unless you explicitly enable a remote model API.
- Shell commands are classified as read/write/search/destructive before
execution. Dangerous patterns (
rm -rf /,dd,mkfs, fork bombs,curl | bash) are blocked. - Destructive operations require explicit user confirmation.
- SQL tool generates queries only; it does not execute them by default.
Destructive SQL (
DROP,DELETE,UPDATE,ALTER) is blocked. - PR Review uses rule-based pattern matching. Results are advisory only and do not substitute for professional security audit.
- API keys are stored locally in
~/.<tool-name>/config.json.
- AI-generated output may contain errors. Users must review before acting.
- PR Review uses pattern matching; it will not catch all bugs or vulnerabilities.
- SQL quality depends on the LLM model and schema information provided.
- Image generation requires an OpenAI API key with DALL-E 3 access. API costs are billed by OpenAI.
- License activation requires internet access for Etherscan API queries.
- This project is maintained by a solo developer. Response times may vary.
See DISCLAIMER.md for full legal disclaimers.
Q: Do I need an API key?
A: Yes. Each tool requires an OpenAI API key (or compatible provider). Set it
via ai --config or the OPENAI_API_KEY environment variable.
Q: What if I send USDT on the wrong network? A: Only ERC20 is supported. Transactions on other networks (BSC, Solana, TRC20) may be lost.
Q: Is my payment refundable? A: All blockchain transactions are irreversible. Please verify the address and network before sending.
Q: What happens after I pay?
A: Run aitools claim <tx_hash>. The CLI verifies the transaction on
Etherscan and activates Pro locally.
Q: Can I use the tools offline? A: LLM features require internet. Shell command classification and basic error diagnosis work offline.
Q: Do you collect my data? A: No. All tools run locally. No telemetry, no analytics, no tracking.
git clone https://github.com/autogz/ai-tools.git
cd ai-tools
# Each tool has its own directory with a venvMIT License. See LICENSE.
Report vulnerabilities via SECURITY.md.
See CODE_OF_CONDUCT.md.