firmis list — List Detected Platforms

You’ve cloned a repo, inherited a project, or just installed a new tool. Before you scan anything, firmis list answers one question: what platforms are even here?

One line. Instant answer.

When to use this

Quick orientation: Before running scan or discover, confirm which platforms Firmis will detect
Checking coverage: You added a new MCP server — run list to verify Firmis picks it up
Scripting and automation: Use --json to programmatically check which platforms are present before conditionally running platform-specific scans
Debugging a scan: If scan seems to be skipping something, run list first — if the platform isn’t listed, Firmis didn’t detect it

For the full picture — component names, file paths, dependencies, and model references — use firmis discover instead.

Usage

firmis list [options]

Example output

Terminal output

Firmis — Detected Platforms
Scanning: /Users/me/my-agent-project

  claude     2 components
  mcp        4 components
  cursor     1 component

3 platforms detected

JSON output (`--json`)

[
  { "platform": "claude",  "components": 2 },
  { "platform": "mcp",     "components": 4 },
  { "platform": "cursor",  "components": 1 }
]

What each detected platform means for your security posture

Platform	What it can access	Key threats
`claude`	CLAUDE.md, tool definitions, permission scopes	Tool poisoning, overgrant permissions, hidden instructions
`mcp`	Tool handlers, transport config, server manifests	Prompt injection, data exfiltration, path traversal
`cursor`	.cursorrules, workspace settings	Rule hijacking, exfiltration via completions
`codex`	Plugin manifests, tool definitions	Supply chain, malicious plugin logic
`crewai`	Agent configs, task chains	Agent-to-agent injection, tool misuse
`autogpt`	Plugin manifests, command handlers	Unauthorized command execution
`openclaw`	Skill definitions, skill handlers	Credential harvesting, network abuse
`nanobot`	Plugin configs, tool handlers	File system abuse, insecure config

The more platforms in your project, the wider your attack surface. A project with claude + mcp + cursor has three distinct threat surfaces, each with their own detection rules.

Options

Flag	Type	Default	Description
`--json`	boolean	`false`	Output as JSON array — useful for scripting or piping into other tools

Examples

List detected platforms

npx firmis list

JSON output for scripting

npx firmis list --json

Conditionally scan only if MCP is present

npx firmis list --json | jq -e '.[] | select(.platform == "mcp")' && \
  npx firmis scan --platform mcp --fail-on high

Platforms — what each platform is and how Firmis detects it
discover — go deeper: component names, file paths, dependencies, and model references
scan — security scan everything that list finds