Secure AI Coding Assistant
The AI that senses what your code needs - before you ask. Spec-driven planning, safe code execution, proxy-backed model access, and multi-agent orchestration - all in one beautiful TUI. Zero API keys. Enterprise-grade security.
The Only CLI That Unifies Security · Intelligence · Speed
Other AI coding tools expose your API keys, stuff MCP descriptions into prompts, and lack real planning workflows. SensAI is the first terminal assistant to converge enterprise-grade security, multi-agent intelligence, and spec-driven planning into a single binary.
All LLM traffic through an audited proxy. Zero raw API keys. OAuth + OS keyring. Pre-flight secrets scanning on every request.
Conductor + sub-agents (Architect, Coder, Tester, Reviewer) with parallel collaboration, 9 LSP tools, and AST structural search.
Cache-first catalog, lazy agent building, cached models, and concurrent splash - warm launches in under a second.
Why Choose
Core capabilities that give developers an unfair advantage
Audit-Grade Reasoning
Security Mode · Sense Protocol
Authorized defensive audits under the Sense Protocol. 7-phase EARS flow, read-only by default, evidence-graded findings, threat-intel correlation, and SARIF/Mermaid dossiers. Paid add-on for eligible Sense Pro subscribers.
Exposed API Keys
Zero-Key Security
All LLM traffic routed through an audited proxy. OAuth + OS keyring authentication. Pre-flight secrets scanning catches leaks before they leave your machine.
Concurrent Agents
Multi-Agent Teams
Conductor orchestrates Architect, Coder, Tester, and Reviewer sub-agents in parallel. Create custom agents from markdown with AI-powered generation.
Phase Workflow
Spec-Driven Planning
Requirements → Design → Tasks with strict approval gates, auto-generated Mermaid diagrams, and TDD plans. Run all tasks or execute manually one by one.
LSP Tools
LSP + AST Intelligence
Semantic navigation with definition, references, hover, rename, code actions, and more. Plus ast-grep structural search across 18 languages.
Languages
Auto-Formatter
Automatic formatting after every agent write with a 3-step fallback: custom → LSP → 25+ built-in CLI formatters including Prettier, Black, gofumpt, and more.
Undo Support
Conversation Checkpoints
Automatic per-turn snapshots of file changes. Restore to any point in the conversation, reverting all modifications and truncating history atomically.
Tool Overhead
Chat Mode
Freeform conversation mode with no tool calls. Perfect for brainstorming, architecture discussions, and code review without any execution overhead.
Compression Levels
Token Compression
Lite, full, and ultra compression modes reduce output verbosity and context size. Dramatically lower credit consumption on long sessions.
Importable Skills
Skills & Rules
Discover and manage Skills with enable/disable toggle, GitHub import, and local folder import. Rules management for context rule files with per-session control.
Context Loading
SENSAI.md Priority Chain
Checks SENSAI.md first, then AGENTS.md - only the first found is loaded. DESIGN.md always loads as additive context. Use /create-sensai to generate from existing AI config files.
SensAI
From planning to execution, SensAI gives your AI agent the deepest toolkit of any terminal assistant.
Spec-Driven Plan Mode
Three-phase planning: Requirements (EARS notation) → Design (Mermaid diagrams, TDD) → Tasks (granular, file-by-file). Strict approval gates with diff previews. Run All for automatic execution or Run Manually for step-by-step control.
Safe Analysis Mode
Read-only deep codebase exploration with worktree-style isolation. Investigate safely without any risk of file changes.
Smart MCP Integration
Dynamic capability discovery - not prompt stuffing. Circuit breaker protection, rate limiting, health monitoring, response scanning, and audit logging.
Multi-Agent Orchestration
Conductor coordinates Architect, Coder, Tester, and Reviewer sub-agents with parallel tool calls and credit-guarded execution. Create custom agents from markdown files with AI-powered generation. Up to 4 concurrent agents.
9 LSP Tools
Definition, references, hover, symbols, rename, code actions, formatting, diagnostics, and restart. Delta-aware feedback after every edit.
AST Structural Search
ast-grep powered pattern matching across 18 languages. Find function signatures, error handling patterns, and API usage by code structure - not text.
Conversation Checkpoints
Automatic per-turn snapshots with full undo. Restore to any point - files revert, history truncates. TUI shows [ Restore ] on every message.
Chat Mode
Zero-tools conversation mode for freeform discussion, architecture brainstorming, and code review without execution overhead. Switch instantly between Chat and Code modes.
Token Compression
Three compression levels - lite, full, ultra - reduce output verbosity and context size. Cut credit consumption on long sessions without losing quality.
Skills & Rules Management
Discover and manage Skills with enable/disable toggle, GitHub import, and local folder import. Rules management for context rule files with per-session control. Import from any public GitHub repo.
SENSAI.md Priority Chain
SENSAI.md loads first, then AGENTS.md as fallback. DESIGN.md always loads as additive context. Use /create-sensai to generate from CLAUDE.md, GEMINI.md, or other AI config files.
Available
Powered by xAI Grok and Anthropic Claude through the SensAI proxy. Toggle Sense Mode for extended context windows up to 2M tokens.
Grok Code Fast
xAIGrok 4.1 Fast (Non-Reasoning)
xAIGrok 4.1 Fast
xAIGrok 4 Fast
xAIGrok 4.20 (Non-Reasoning)
xAIGrok 4.20
xAIGrok 4.20 (Multi-Agent)
xAIGrok 4.3
xAIClaude Haiku 4.5
AnthropicClaude Sonnet 4.6
AnthropicClaude Opus 4.6
AnthropicClaude Opus 4.7
AnthropicIntroducing
A specialized high-tier audit mode operating under the Immunisense Sense Protocol v1.0. Authorized defensive vulnerability discovery, risk analysis, and remediation planning inside SensAI's auditable workflow. Available as a paid add-on for eligible Sense Pro subscribers.
Authorized defensive audits. Evidence before severity.
Security Mode swaps the standard coding agent for the Sense Protocol - a reasoning contract trained on authorized vulnerability discovery, threat modeling, and responsible remediation. Every tool call runs under a strict read-only policy by default, with separate approval gates for discovery, verification, remediation, and any fix execution.
● Eligibility: active Sense Pro subscription · verified organization · signed rules of engagement · admin approval.
Read-Only Enforcement
Every tool invocation is validated against the Sense policy before dispatch. Write-class tools are denied at the runtime boundary and only become available if a fix phase is explicitly approved under an active engagement.
Seven non-negotiable rules
The Sense Protocol ships with an enforced behavioral contract. The agent cannot override it - every audit session observes these rules end-to-end.
Authorization first
Confirms scope, targets, exclusions, test intensity, network limits, destructive-check policy, and compliance mappings before audit work begins.
Read-only first
Safe analysis is the default. Remediation diffs and tests are proposed, never written - unless a fix phase is explicitly approved.
Evidence before severity
Every candidate finding is graded Confirmed, Likely, Needs Verification, or Informational. No claim ships without evidence.
Defensive boundaries
No credential theft, stealth, persistence, destructive exploitation, public disclosure, or weaponized chains. Minimal proof only.
Threat intelligence
Correlates findings with CISA KEV, NVD, CWE, MITRE ATT&CK, Exploit-DB, EPSS/CVSS, and known campaign TTPs when tools permit.
Visual reasoning
Produces Mermaid diagrams for attack surface, trust boundaries, data flows, and major attack paths on every audit report.
Approval gates
Discovery, verification, remediation proposal, and fix execution are separated - each gated by explicit user approval.
Seven phases. One signed audit dossier.
Intake to reporting is gated by explicit user approval. No phase advances silently.
Intake
Rules of engagement and requirements captured in EARS notation. Scope, targets, exclusions, destructive-check policy locked in.
Recon
Architecture modeling from safe read-only tools. Assets, trust boundaries, data flows, and system surface enumerated.
Discovery
Systematic sweep across code, config, dependencies, auth, data flow, secrets, permissions, supply chain, and deployment surfaces.
Verification
Chaining analysis and threat-intel correlation. Candidates are verified and promoted to Confirmed or downgraded with evidence.
Remediation
Root cause, remediation proposal, security tests, and validation steps - staged behind an approval gate before any fix is applied.
CVE Review
Novel vulnerability assessment and responsible-disclosure notes when a finding may warrant coordinated disclosure.
Reporting
Final dossier: Markdown narrative, Mermaid diagrams, SARIF-shaped findings, JSON summary, executive dashboard, compliance mapping.
Finding Schema
Every finding is emitted with a consistent shape. Reports include Markdown, Mermaid, SARIF-compatible JSON, executive dashboard, and compliance mapping.
Specialized security roles
Sense orchestrates specialized roles and merges their outputs into a single evidence ledger. Nothing lands in the dossier without provenance.
Simple
Three balance buckets consumed in order: tier → bonus → top-up. Only tier credits reset each billing cycle.
Free
forever free
credits/month
- 50 credits/month
- Grok Code Fast model
- Non-reasoning only
- 256K context window
Pro
/month
credits/month
- 500 credits/month
- All models
- All reasoning levels
- Up to 2M context (Sense)
Ultra
/month
credits/month
- 1,500 credits/month
- All models
- All reasoning levels
- Up to 2M context (Sense)
Sense
/month
credits/month
- 4,000 credits/month
- All models
- All reasoning levels
- Up to 2M context (Sense)
Sense Pro
/month
credits/month
- 10,000 credits/month
- All models
- All reasoning levels
- Up to 2M context (Sense)
- Eligible for Security Mode add-on
Security Mode
Specialized high-tier audit mode powered by the Sense workflow and enforced read-only tool policy. Requires an active Sense Pro subscription.
/month
Billed alongside Sense Pro
Flat rate · no annual discount
● Eligibility: active Sense Pro subscription · verified organization · admin approval required before provisioning.
Ready to Sense Your Code?
Install in one command. Authenticate with your browser. Start coding with the most secure AI assistant ever built for the terminal.