Roadmap

What exists, what is being built, and what follows. Anything below the Shipped section does not exist yet, and is labelled so you never have to guess which is which.

Shipped

Available now. Everything below this line is not.

Analysis engine

Token, wallet, contract and website analysis against live Robinhood Chain data, with every finding carrying the source it came from. Why it matters: A score with no evidence behind it is an opinion. Every number here can be checked at its source, and a check that could not run is reported as unknown rather than passed.

Console

Run an analysis, watch it stream, read the report, compare two subjects, and keep a history scoped to your key. Why it matters: The evidence is complete about a third of the way into a run. Streaming means you read the findings while the summary is still being written instead of waiting for both.

Molthood Agent

A conversational surface that reaches the same analysis engine and answers in plain language, with a model selector and full conversation history. Why it matters: Most questions are not a form. Asking one in a sentence is faster than choosing a subject type, and the Agent decides for itself when live data is needed.

Change detection

Re-run a subject and receive what changed rather than a second full report. Why it matters: A single analysis is a photograph. Risk moves — liquidity drains, ownership transfers — and the difference between two runs is the part worth reading.

Current

Being built. Partly usable, and honest about which parts.

Public API — Coming soon

The same engine the console uses, over plain HTTP and JSON, with a documented response contract. Why it matters: An analysis platform you cannot call from your own code is a website. The reference is written and the routes exist; what remains is the stability guarantee.

API keys — Coming soon

Scoped credentials with per-key allowances, rotation, revocation and an audit trail. Why it matters: Analyses cost real inference credit, so they have to be metered against something. Per-key limits also mean one integration cannot spend another's budget.

Next

Starts when the current work lands. Shape is decided.

Webhooks — Coming soon

Delivery of execution and change events to an endpoint you control, with signing and retries. Why it matters: Polling for a result that takes a minute wastes both sides. A monitored subject that changes at 3am should reach your system without anyone watching a screen.

Notifications — Coming soon

In-product and email alerts for watched subjects, scoped to what actually changed. Why it matters: A watch is only useful if it can reach you. The diff already exists; this is the delivery half of it.

SDK — Coming soon

Typed clients for TypeScript, Python and Go. Why it matters: Types that match the response contract mean a field that moves breaks at compile time rather than in production at the worst possible moment.

Planned

Committed, not started. Shape is roughly known.

CLI — Coming soon

Drive executions, stream them, and pull artifacts from a terminal. Why it matters: Scripting an analysis into a build step or a cron job should not require writing an HTTP client first.

MCP support — Coming soon

A Model Context Protocol server exposing Molthood's analyses as tools any compatible AI client can call. Why it matters: Your assistant should be able to perform the analysis and read the evidence, rather than describe what it would check if it could.

Skills — Coming soon

Named, versioned workflows: package a sequence you run often, install one somebody else built. Why it matters: Most analysis is repetition with the subject swapped. A skill makes that repetition shareable instead of rebuilt each time.

Portfolio — Coming soon

Every holding in a wallet screened by the same rules a single token gets, with the weakest positions surfaced first. Why it matters: Risk concentrates in the holding nobody checked. A portfolio score is a ceiling when a check could not run — never a clean bill of health by omission.

Smart alerts — Coming soon

Conditional monitoring: tell it what would matter, and hear only about that. Why it matters: An hourly report about a token that has not moved is noise you learn to ignore, which defeats the point of watching at all.

Future

Wanted, not committed. Listed so the direction is legible.

Wallet intelligence — Coming soon

Behavioural profiling of an address: what it does, how it has behaved over time, and how it relates to others. Why it matters: A balance says what an address holds. What it has repeatedly done says considerably more about what it is.

Agent marketplace — Coming soon

Publish and install specialist agents built by other people, versioned and reviewable. Why it matters: The set of useful checks is larger than one team can write. The engine is already agent-shaped; the gap is distribution.

Strategy builder — Coming soon

Compose conditions and analyses into a saved, re-runnable strategy without writing code. Why it matters: The people with the best sense of what to check are frequently not the people who want to write a client for it.

Mobile app — Coming soon

Native iOS and Android clients for alerts and quick checks. Why it matters: Checking a contract before signing is something you do standing up, on a phone, with a wallet already open.

Browser extension — Coming soon

Inline risk context on the pages where a token address actually appears. Why it matters: The decision happens on the swap screen, not in a separate tab. Evidence that arrives after the decision is history, not risk analysis.

How to read this

A feature moves down this page, never up. Something in Future is wanted rather than committed, and listing it is a statement of direction — not a queue position. If something you need is missing, the roadmap is the wrong place to find that out; say so instead.