FAQ
Scope, engine behavior, privacy, and the boundaries of the current release.
01Does Signalint replace my linter or TypeScript?
No. Oxlint, TypeScript, and optional Biome remain the diagnostic sources. Signalint invokes them, normalizes their output, caches safe reuse, and clusters the response for an MCP client.
02Does project code leave my machine?
Signalint is a local stdio server and does not add a network transport. The MCP client using the response has its own separate data-handling policy.
03Which languages are supported?
JavaScript and TypeScript projects only. The built-in engines are Oxlint, tsc, and Biome; arbitrary custom engines are not supported in v1.
04Does Signalint apply fixes?
No. It reports fixable: true only when an engine supplies a structured fix, but v1 does not modify source files.
05Is Signalint a security scanner?
No. It is not a SAST tool and does not claim to find vulnerabilities. Signalint’s own MCP argument boundary is documented in the repository threat model.
06How do monorepos work?
The tsc adapter requires one tsconfig.json at the project root. Monorepos should provide a solution-style root config using TypeScript Project References.
07What does loop detection track?
Only normalized lint, type, and test issue signatures that disappear and return repeatedly. It does not classify the agent’s general conversation.
08Is there an IDE extension?
Not yet. Current integrations use MCP clients or the command-line check and stats commands.
09What does it cost?
Signalint is free and open source under the MIT License. It runs locally and has no hosted pricing tier.