About Trie Memory
Trie Memory is an MCP server published by tomas-samek in the Search category: experimental MCP memory server pairing a delta-encoded recognition trie with a concept store that binds surface forms across languages via temporal co-occurrence — "bush", "křoví", "茂み" collapse into one concept without a dictionary. Append-only commit layers record what was learned and when.
Recent commit activity is the strongest signal that reported issues will get attention.
Install
git clone https://github.com/tomas-samek/trie-memoryThis server has no ConduID identity, so agent calls to it are not receipted. Pin the version you install and review the source before granting it credentials.
Tools (6)
trie_path_keyContent-addressable path key (byte or word trie).
Parameters not published.
trie_queryRead-only probe; returns `deepest_node`, `match_depth`, `depth_profile`.
Parameters not published.
trie_readWalk a leaf → root and return node summaries.
Parameters not published.
trie_statsPer-node: visit count, depth, children count, spectrum, state.
Parameters not published.
trie_tokenizeShow how the word tokenizer splits + assigns silence gaps.
Parameters not published.
trie_writeFeed a string into both tries.
Parameters not published.
Ask AI
Ask AI about Trie Memory
Powered by Claude · Grounded in docs
Security checks
6 checks have not been run yet.