hive-proximum.store — io.github.hive-agi/hive-proximum 0.1.18

Datahike-backed IMemoryStore + IMemoryStoreTemporal for hive-proximum.

Bitemporal design:

• Every write is a new Datahike transaction (immutable fact log)
• Updates add new facts; old values preserved in history
• Deletes set :entry/deleted? true (tombstone, never hard-delete)
• Temporal queries use d/as-of and d/history for time-travel

Protocol extension uses extend (function, not macro) to avoid compile-time dependency on hive-mcp.protocols.memory.

Backend: Datahike 0.8.1667 (matches hive-mcp/hive-knowledge)

create-store clj

(create-store)

(create-store opts)

Create a ProximumStore instance. Does NOT connect immediately (lazy). Options: :backend - :memory (default) or :file :path - file path for :file backend (default: data/proximum/datahike) :config - raw Datahike config map (overrides :backend/:path)

create-stub-store clj

(create-stub-store)

Create a ProximumStore. Replaces the old stub. Uses in-memory Datahike backend by default.

ensure-conn! clj

(ensure-conn! {:keys [conn-atom cfg]})

Ensure Datahike connection is established. Returns conn.

entity->map clj

(entity->map e)

Convert Datahike entity map to clean entry map.

:entry/content is stored as a JSON-encoded string by add-entry! / update-entry! (see entry->tx-data). Parse it back to a Clojure map on read so callers (e.g. kanban transitions) see structured content without round-tripping through json/read-str. Non-JSON content (legacy plain strings) is returned as-is.

entry->tx-data clj

(entry->tx-data {:keys [id type content tags content-hash source expires-at]})

Convert entry map to Datahike transaction data.

:entry/content is canonically JSON-encoded so entity->map on read can deserialize back to a Clojure map. Pre-encoded strings (e.g. JSON produced by memory-kanban) pass through untouched.

entry-schema clj

extend-store-protocols! clj

(extend-store-protocols!)

Extend ProximumStore with IMemoryStore + IMemoryStoreTemporal. Uses extend (function) — protocol resolution is fully runtime. Safe to call multiple times (idempotent via atom guard).

make-datahike-config clj

(make-datahike-config)

(make-datahike-config {:keys [backend path]
                       :or {backend :memory path "data/proximum/datahike"}})

Build Datahike config map. Supports :memory and :file backends.

pull-entry clj

(pull-entry db id)

Pull an entry from db, return nil if deleted or not found.

pull-entry-raw clj

(pull-entry-raw db id)

Pull an entry from db, including deleted entries.