Manamancer is a free deck builder for Commander and thirteen other Magic formats. Search the current card index, assemble a list with live format checks, inspect its mana curve and color requirements, test opening hands, track the cards you own, and publish a deck page that other players can read or clone.
Deckbuilding information, not just a card list
Every deck is evaluated against the rules of its selected format while you work. The statistics view turns the list into decisions you can act on: mana value distribution, colored pips, land and mana-source estimates, and practical counts for ramp, card draw, removal, board wipes, and the other jobs a deck needs filled. Sideboards and maybeboards stay separate from the main-deck total, and Commander-specific zones such as commanders, partners, and companions are represented explicitly.
Manamancer supports Commander, Standard, Pioneer, Modern, Legacy, Vintage, Pauper, Alchemy, Historic, Timeless, Standard Brawl, Brawl, Pauper Commander, and Canadian Highlander. The builder applies the relevant deck size, copy limit, sideboard, color identity, and legality rules instead of treating every format like Commander.
What the numbers are actually telling you
A hundred-card singleton deck fails in predictable ways, and most of them show up in the statistics before they show up at the table.
The most common is a mana base that does not match the curve sitting on top of it. A deck whose average mana value climbs past three needs more sources than one built around two-drops, and the count that matters is not "lands" but lands plus mana rocks plus mana dorks plus rituals. Manamancer separates those, because thirty-six lands and two rocks plays nothing like thirty-two lands and six rocks even though both read as thirty-eight sources.
Color requirements are the second. Counting how many cards of each color you run tells you very little; counting pips tells you what your lands have to produce. A three-color deck with a double-black card on turn three has a real constraint that a flat color count hides completely. The pip breakdown is there so you can see which color is quietly demanding the most from your mana base, and fix it before the deck teaches you the hard way.
The third is role coverage. Commander decks tend to be built card by card, and it is easy to finish a list that is short on interaction or has no way to refill a hand. The role counts are heuristics read from Oracle text and card types, so they are a prompt rather than a verdict: a card that draws two and makes a creature counts once, and a flexible card may fill a role the text pattern misses. Treat a low number as a question worth asking rather than a problem you must fix. How to read your deck statistics covers what each figure counts and where the estimates stop being reliable.
Format checks that reflect the format
Legality is not one rule. A Commander deck is a hundred singleton cards inside a color identity; a Modern deck is sixty-plus with a four-copy limit and a fifteen-card sideboard; Pauper restricts the pool to commons; Canadian Highlander applies a points list rather than a ban list. Checking a list against the wrong set of rules is worse than not checking it at all, because it looks like an answer.
Manamancer keeps those differences in a dedicated format engine, so switching a deck's format re-evaluates it against the rules that actually apply: deck size, copy limits, sideboard and command-zone rules, color identity, and the current banned or restricted list. Colour identity in particular catches people out, because it counts mana symbols in rules text, not just the casting cost. Understanding format legality explains what the checks can and cannot tell you, and why an event's official rules remain the final authority.
Search backed by real card data
Card search uses a local index derived from Scryfall data and understands familiar Scryfall-style filters. Results include Oracle text, types, color identity, format legality, printings, and market estimates. New card data is refreshed regularly, and exact printing choices can be preserved when a deck or collection is saved. Card names and legality are data-backed throughout the app rather than generated from free text.
A Commander assistant with a hard validation boundary
Any signed-in member can ask the in-app deck agent to review a Commander list — AI is metered by a monthly allowance rather than reserved for a paid plan. It receives the same computed statistics shown in the interface, searches the local card index, and proposes additions or cuts for the player to approve. Before a recommendation is displayed, the server verifies that the card exists, is Commander-legal, and fits the commander's color identity. Suggestions that fail those checks are discarded rather than shown.
That boundary is narrow on purpose: it prevents invented card names from becoming actionable advice. It does not make the model a rules judge, and every proposal is still yours to apply or dismiss. How the validated Commander Agent works describes exactly where the guarantee starts and stops.
Collection, playtesting, and sharing
A personal collection can record printing, finish, condition, binder placement, and quantities. The deck builder can show which cards you already own without exposing your collection to other players. The playtest table draws and shuffles the real saved list, so an opening-hand check reflects the deck you actually built.
Public deck pages can include a primer, card categories, deck tags, full statistics, comments, likes, and a link another player can use to clone the list. You can browse what other people have published on the Explore page, or read more about the project on the about page.
Manamancer is unofficial fan content. Magic: The Gathering and its card imagery are owned by Wizards of the Coast. Card data and images are provided by Scryfall.