What it does
The local Agent Core MCP accepts agent-generated resource history, checks renderer-critical shapes and privacy safeguards, builds a teacher-review preview, and exports a .allopack.json payload. It does not silently call a model, read API keys, upload source text, publish content, or write to arbitrary paths. The publisher's documentation also covers the remediation pipeline, evidence boundaries, a portable Skill and connector status.
It suits educators, special education teams and school psychologists who want agents to draft accessible, differentiated resource packs without opening the AlloFlow interface. Teachers review each preview before anything is shared.
The publisher describes three ways to run AlloFlow: the browser-based Gemini Canvas path, which needs no installation, AlloFlow Desktop for a local-first teacher laptop, and an optional Docker School Box for server or appliance use. Opening an original text does not require an AI call, but generating explanations does.
Limitations and requirements
The Gemini Canvas path requires an eligible Google account, and availability and school administrator settings apply. Cloud-provider terms, data handling and costs depend on the deployment chosen. Generated source material should be reviewed before teaching from it.
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
AlloFlow Document Remediation
Local-first document accessibility remediation using the canonical AlloFlow app pipeline.
v0.3.5As shown on registry.modelcontextprotocol.io when we last read it