Mitsu CLI
Mitsu CLI uses Guided mode to ask the operator for a decision before taking the next step in an authorized security assessment.
- 539 ★ on the repository
- Last commit
- MIT
Broad access
Why this level
- The dual-use tool runs checks and commands; AutoPwn can affect the target system.
Install
In your terminal, with SkillFoxx CLI
npx skillfoxx add cli/mitsu-cliDetects the agents on your machine, shows the access level and pins the version.
Other ways to install
Access and checks
How we reviewBroad accessThe tool can change data, run commands or handle secrets.
- The dual-use tool runs checks and commands; AutoPwn can affect the target system.
This is third-party code. Review the repository files before installing.
What it does
Mitsu CLI organizes a security-assessment session with a chosen model. It retains targets, services, observations, findings and evidence. The agent chooses its next step from that information and returns results to the same session. An internal MCP layer connects selected utilities, external servers and user-defined commands. Recon limits work to passive research, Guided asks the operator to steer, and AutoPwn allows more independent action. Session material can produce an attack-surface map and assessment report.
Who it is for. Security assessors with an agreed scope who need a shared session and evidence for a report.
Good fit when
- A specialist collects information about an authorized test environment.
- A finding is linked to evidence in the session.
- An operator chooses the next investigation direction in Guided mode.
- After assessment, the team prepares a report and marks unverified observations.
Not a fit when
- The target owner has not authorized testing.
- Targets and permitted actions cannot be restricted.
- Automatic mode would run against production without agreement.
Example request
Create an authorized lab session in Recon and show its scope without active actions.Limitations
Python 3.10 or later, your own model provider and separately available utilities are required. Automatic mode can perform active actions and change a target system; define written scope and start with passive Recon. Tools run with operator permissions, so restrict the account and its access to files. Sessions retain target, account and evidence details; treat them as sensitive information and restrict access to the saved material.
How to disable. Stop the active session and disable selected tools and external MCP servers. Clean retained material and credentials separately.
FAQ
How do I change modes?
Enter `/mode` in the session. Recon keeps work passive, Guided waits for operator decisions, and AutoPwn permits independent active actions. Select a mode within the agreed assessment scope.
How do I change the model?
Use `/connect` to configure a provider, then `/model` to choose an available model. For Ollama, first start the local server with a model installed.
Where are tools configured?
Open `/mcp` to select tools and external servers. The full configuration is stored in `~/.config/mitsu/mcp_servers.json`; keep only tools approved for the assessment before connecting.
Are missing utilities installed automatically?
The program can propose installation and request user approval.
Can I use a local model?
Yes. A separately running Ollama is supported.
Discussion
No questions or reviews yet.
Related
CLI toolsA CLI for auditing agent MCP servers: inventory, evidence per server and a CI policy where unmeasured never counts as clean
An MCP memory server for agents plus a free gate that stops dangerous commands before they run
A CLI agent for authorized pentesting: from a natural-language brief it runs recon, finds and verifies vulnerabilities and prepares a report
Security Detections MCP: a detection rule database
Security Detections MCP
An MCP server that merges Sigma, Splunk, Elastic and KQL rules into one database for finding detections and ATT&CK gaps
MCP checks shell commands, while shellfirm hooks apply rules to supported agent calls before they execute
CLI mediator scans agent network requests and MCP calls for secret leaks, unsafe destinations and injected instructions
Comparison with similar tools
Cards that share the main search query, and how they differ.
| Tool | Official | Authentication | Repository stars | Main difference |
|---|---|---|---|---|
| Mitsu CLI (this card) | no | 539 | ||
| Strix | no | 75,080 | performs assessments with agents in Docker; Mitsu provides a session CLI with mode and utility-catalog selection | |
| AdStrike | no | not required | 455 | An MCP server gives agents Active Directory assessment tools and shared engagement context for authorized security tests |
README in short
The internal MCP layer serves CLI tools. Users configure external servers, and separate sessions retain material from different assessments.