Fortify automatically scans workspace signature files, connects to OpenAI, Anthropic Claude, Google Gemini, and offline Ollama models, and offers an interactive `$EDITOR` workflow directly in your terminal.
███████╗██████╗ ██████╗ ████████╗██████╗ ███████╗██╗ ██╗ ██╔════╝██╔═══██╗██╔══██╗╚══██╔══╝╚═██╔═╝██╔════╝╚██╗ ██╔╝ █████╗ ██║ ██║██████╔╝ ██║ ██║ █████╗ ╚████╔╝ ██╔══╝ ██║ ██║██╔══██╗ ██║ ██║ ██╔══╝ ╚██╔╝ ██║ ╚██████╔╝██║ ██║ ██║ ██████╗██║ ██║ ╚═╝ ╚═════╝ ╚═╝ ╚═╝ ╚═╝ ╚═════╝╚═╝ ╚═╝
───────────────── Fortify v0.6.0 Chat ─────────────────
[SUCCESS] Loaded project context: fortify (Node.js)
You > Please audit @src/index.js and check @security-check
📎 Loaded file: src/index.js (774 bytes)
Assistant > I have audited src/index.js against security guidelines. No exposed secrets detected.
──────────────── Authentication & Provider Setup ────────────────
Select active AI Provider to configure:
1) OpenAI (GPT-5.4, GPT-5.3)
2) Anthropic (Claude 3.5 Sonnet)
3) Google Gemini (Gemini 2.0 Flash / 1.5 Pro - Free Tier Available)
4) Ollama (Local LLM)
Choice [1-4]: 3
[SUCCESS] Google Gemini API key saved and active provider set to 'gemini' in ~/.fortify/config.json
──────────────── Interactive Commit Assistant ────────────────
Branch: main | Style: conventional | Scope: cli
Diff Summary: 3 files changed, 45 insertions(+), 2 deletions(-)
Generating commit message suggestion...
Opening $EDITOR (code --wait) for interactive fine-tuning...
[SUCCESS] Conventional Commits format validation passed!
Committed: "feat(cli): add Google Gemini multi-provider support"
──────────────── Project Initialization ────────────────
Scanning directory for framework signatures...
✔ Detected project stack: Node.js (package.json detected)
✔ Created workspace config: .fortify/project.json
✔ Updated .gitignore to exclude local credentials
Your workspace is now fortified! Run `fortify chat` to start.
Switch seamlessly between cloud leaders or run 100% offline local LLMs.
Full support for GPT-5.4, GPT-5.3, and lightweight fallback models with streaming responses.
Native Claude 3.5 Sonnet & Claude 3 Opus streaming via Anthropic Messages REST API.
Gemini 2.0 Flash & 1.5 Pro integration with generous free tier token quotas from Google AI Studio.
100% offline local LLM execution (Llama 3, DeepSeek, Mistral) over local HTTP port 11434.
Compare Fortify against existing developer CLI assistants.
| Feature | 🛡️ Fortify CLI | GitHub Copilot CLI | Aider | Cursor CLI |
|---|---|---|---|---|
| Multi-Provider (OpenAI, Anthropic, Gemini, Ollama) | ✅ Native (`-p, --provider`) | ❌ (OpenAI only) | ✅ | ❌ (Proprietary) |
| Repo Stack Auto-Detection (`fortify init`) | ✅ Built-in | ❌ | ❌ | ⚠️ Basic |
| Interactive `$EDITOR` Commit Workflow | ✅ Built-in (`-i`) | ❌ | ❌ | ❌ |
| Workspace Prompt Shortcuts (`.fortify/plugins`) | ✅ Native | ❌ | ❌ | ❌ |
| Smart `@file` Reference Autocomplete | ✅ Native | ❌ | ✅ | ❌ |
| Conventional Commits Format Validator | ✅ Native (`--validate`) | ❌ | ❌ | ❌ |
Powerful terminal triggers designed for developer speed.
fortify auth
Interactive credential setup for OpenAI, Anthropic, Gemini, or Ollama.
fortify chat -p gemini
Start interactive session overriding AI provider or model on the fly.
fortify commit -i --validate
Draft commit messages, edit in `$EDITOR`, and validate format before committing.
fortify explain ./logs/error.txt
Analyze stack traces, runtime errors, or inline code snippets.
fortify summarize src/
Recursively summarize codebase directory structure with token safety caps.
fortify plugin list
Inspect local workspace prompt shortcuts (`@security-check`, `@refactor`).