Road Poneglyph plan review server

Road Poneglyph is a small helper that makes your AI look at a plan and tell you what is missing from it. Instead of praising what you wrote, it hunts for the things you never said out loud. It is handy for anyone who has to hand a plan to other people, like project leads, product managers and team coordinators.

What this MCP server does

You paste or point your AI at a plan, and you ask it to read the poneglyph. The AI passes your plan to this helper, which hands back a carefully built set of review instructions. Your AI then follows those instructions and writes the analysis itself. You get back three short sections about what your plan quietly assumes, what technical risks it skips, and what could go wrong while you actually try to ship it.

Click to zoom

What you can do with it

• Spot assumptions your plan treats as facts without proof
• List technical and architecture risks the plan does not mention
• Point out operational and execution blind spots
• Review plans for software projects or other kinds of projects
• Run the same review on a plan again after you fix things
• Get a short structured critique instead of a vague opinion

Try asking your AI

• “Read the poneglyph on this launch plan and tell me what I left out.”
• “Here is our migration plan. Run a pre-mortem and list the hidden risks.”
• “Review this product plan as a Marketing project and show the gaps.”
• “What assumptions is this plan quietly relying on?”

What it gives back to you

You get a written analysis in your chat, split into three parts: unstated assumptions, omitted technical risks, and execution blind spots. Each part is a short list of points your AI pulled from your plan. Nothing is saved anywhere and no files are created, it is just text in the conversation.

Before you start

Who it's for

Project leads, product managers, founders and anyone who writes plans and wants a second pair of eyes before sharing them.