MCP servers · blackcat-informatics

PurRAssist: recovery of intent from imperfect AI output

An MCP server, Rust crate and CLI that recovers intent from imperfect JSON, tool calls, names, paths and Markdown edits, and reports what changed.

This text was machine-translated from the publisher's own information and has not been checked by a human editor. The publisher's original text is the reference.

This resource was imported automatically from the sources listed below and has not been reviewed by the AI Gate team yet. The description is the publisher's own, as published. Check the repository before installing.

What it does

PurRAssist takes a schema and an effort level, searches for composed corrections to imperfect JSON, names, paths and text, validates the complete result, and returns evidence of what changed. Ambiguous alternatives stay visible, and a schema-valid result is still treated as a guess. It also covers source-bound text edits, Markdown editing, tool-call recovery from raw model output across many dialects, and grammar inspection for Python, Rust, JavaScript/JSX and TypeScript/TSX.

It suits cases where an AI's output is almost correct: malformed tool calls, JSON with wrong keys or enum values, mistyped paths, or Markdown edits that need to be applied safely. The same engines and evidence receipts are available as a Rust crate, an MCP server or a command-line tool, so the same recovery can be called from code, from an MCP client or from a shell.

The MCP server speaks newline-delimited JSON-RPC over stdio and can be added to any MCP client that launches stdio servers. Starting it creates a .purrassist/ directory in the root, where file tools keep transaction journals. The CLI runs a tool with a root directory and a JSON argument object read from stdin, and returns the same results as MCP. The release is version 0.1.0, with six crates published to crates.io.

Limitations and requirements

File tools need a root directory, and the server creates a .purrassist/ directory there. Grammar acceptance does not prove program correctness. A schema-valid result is still a guess, and ambiguous alternatives remain visible rather than resolved. Tool-call markup in a dialect that is not built in, or not on the request's allow-list, is rejected as unsupported. Plain-number support, graph loading and execution are planned, not available. Cold grammar compilation, higher-level emitter preparation and full seeded or witness generation are outside the current operation. The publisher notes known limitations, which it says are not release blockers.

Safety

Our team has not reviewed this resource. Only automated rules checked it; resources with high-risk signals are not published. These are signals, not a guarantee that installing it is safe. Try it in a sandbox project first.

  • Runs code or changes an assistant's behaviour

Sources

  • Official MCP Registry first seen 13 Nov 2026 · last seen 13 Nov 2026 · its page on the source
    PurRAssist

    Recovers intent from imperfect AI output: JSON, tool calls, names, paths and Markdown edits.

    v0.1.0
    As shown on registry.modelcontextprotocol.io when we last read it
MCP servers
neondatabase

Neon MCP Server (Postgres)

Neon's MCP server for managing Neon Postgres in plain language: create projects and branches, run queries and prepare schema migrations from your editor.

663 GitHub starsData & databasesSoftware development
Updated 17 Nov 2026
MCP servers
supabase-community

Supabase MCP Server

Supabase's MCP server that connects an AI assistant to your Supabase project to design tables, run queries, read configuration and look up documentation.

Data & databasesSoftware development
Updated 17 Nov 2026
MCP servers
hashicorp

Terraform MCP Server (HashiCorp)

HashiCorp's MCP server that gives an assistant current Terraform Registry information on providers, modules and policies, and can manage HCP Terraform workspaces.

1.6K GitHub starsCloud & DevOpsSoftware development
Updated 17 Nov 2026
MCP servers
modelcontextprotocol

MCP reference servers

Seven small example servers maintained by the Model Context Protocol project: files, Git, web fetching, memory, time, step-by-step thinking and a test server.

92K GitHub starsSoftware development
Updated 17 Nov 2026