
Satori Aligner
Scan LLM inputs and outputs for factual misalignment and semantic bias before you ship agent features to users.
Overview
Satori-Aligner is a MCP server for the Ship phase that scans LLM inputs for factual alignment issues and semantic bias.
What is this MCP server?
- Satori-Aligner remote MCP (SSE) at satori-aligner-mcp.vercel.app
- Factual alignment checks on LLM-bound content
- Semantic bias scanner for prompts and retrieved context
- Premium access via optional EIP-3009 payment-signature header
- Shipped as version 1.0.0 MCP server schema entry
- Version 1.0.0 in MCP server.schema.json metadata
- One SSE remote at satori-aligner-mcp.vercel.app/api/mcp
- Product title Satori-Aligner in catalog
What problem does it solve?
Agent prompts and RAG chunks often ship with hidden bias or unsupported facts that only show up in production complaints.
Who is it for?
Builders adding LLM features who need a lightweight MCP review pass before launch without building an in-house eval stack.
Skip if: Regulated clinical or legal deployments that require formal human compliance sign-off and documented audit trails only this stub does not describe.
What do I get? / Deliverables
You catch alignment and bias risks during review so releases go out with cleaner prompts and safer model-facing text.
- Alignment findings on supplied LLM inputs
- Semantic bias signals for prompt and context review
- Actionable review notes for ship checklists
Recommended MCP Servers
Journey fit
Alignment checks belong in Ship when you review prompts, RAG context, and model-facing copy for safety and truthfulness. Bias and factual alignment review is a pre-release quality gate, adjacent to code review but focused on model behavior.
How it compares
LLM alignment MCP scanner, not a full red-team pentest suite or a generic linter skill.
Common Questions / FAQ
Who is Satori Aligner for?
Solo developers and small teams shipping LLM-powered products who want MCP-accessible factual and bias checks on inputs.
When should I use Satori Aligner?
During Ship review when you freeze prompts, system instructions, or retrieved context ahead of a user-facing release.
How do I add Satori Aligner to my agent?
Configure the remote SSE URL https://satori-aligner-mcp.vercel.app/api/mcp in your MCP client and include payment-signature if premium settlement applies.