Reference
Frequently Asked Questions
Answers to common questions about SPECTRALE discovery, external components, MCP access, and multi-project use.
Does SPECTRALE need access to all my infrastructure?
No. Discovery operates using the information available through supported sources.
The resulting architecture model depends on what can be observed or inferred from that evidence.
Does detecting an external service mean SPECTRALE can access its data?
No.
External service discovery and authenticated third-party access are different capabilities.
A detected service may appear in the architecture without an active connection to that service.
Does SPECTRALE replace my existing development tools?
No.
SPECTRALE is intended to complement existing development environments by providing a structured view of the broader architecture.
Can AI agents modify my infrastructure through MCP?
Not through the current read-only SPECTRALE MCP toolset.
The exposed tools focus on retrieving architectural information.
Is SPECTRALE limited to one AI client?
No.
The MCP interface is designed around an open protocol.
Actual compatibility depends on the client's supported transport and authentication methods.
Will SPECTRALE discover every component automatically?
Not necessarily.
Discovery results depend on available evidence, supported sources, and the visibility of the underlying system.
Can I use SPECTRALE across multiple projects?
Yes.
SPECTRALE supports separate project contexts, allowing architectural information to be organized independently.
Why use MCP instead of copying architecture documentation into an AI chat?
MCP gives compatible AI clients a structured way to request relevant information from SPECTRALE.
Instead of relying only on manually copied documentation, an agent can request specific project, component, and dependency information when needed.
This can reduce manual context preparation, although the quality of the resulting assistance still depends on the available data and the AI client.