Arcticdb MCP

Findings (2)

• lowLow. Minor, or common in normal tools. Shown for information.
  No manifest declaring network access and environment variables (agent.json)
• lowLow. Minor, or common in normal tools. Shown for information.
  Contacted these domains while starting up (no manifest to compare against)

  pypi.org

Network activity

Tools exposed (49)

• get_uri

  Return the ArcticDB URI used by this MCP server connection.

• describe

  Return a compact summary of the ArcticDB store. If library is provided, describes only that library. If omitted, describes all libraries. Returns per symbol: row count, column names, date range, latest version number, last update time, and metadata. Useful as a first call to orient Claude before a

• modify_library_option

  Modify a configurable library option. Supported options include DEDUP, ROWS_PER_SEGMENT, COLUMNS_PER_SEGMENT, REPLICATION, and BACKGROUND_DELETION.

• write_batch

  Write multiple symbols in one call. payload format: {"symbol": str, "data": list[dict], "metadata"?: any, "index_column"?: str}

• append_batch

  Append to multiple symbols in one call. payload format: {"symbol": str, "data": list[dict], "metadata"?: any, "index_column"?: str}

• update_batch

  Update multiple symbols in one call. payload format: {"symbol": str, "data": list[dict], "metadata"?: any, "date_range"?: [start, end], "index_column"?: str}

• delete_batch

  Delete multiple symbols or symbol versions in one call. request item can be symbol string or {"symbol": str, "version_ids": [int, ...]}.

• read_batch

  Read multiple symbols in one call. symbols items can be symbol strings or request dicts with: symbol, as_of, date_range, row_range, columns, output_format.

• read_metadata_batch

  Read metadata for multiple symbols in one call. symbols items can be symbol strings or request dicts with symbol/as_of.

• get_description_batch

  Get schema/description for multiple symbols in one call. symbols items can be symbol strings or request dicts with symbol/as_of.

• write_metadata_batch

  Write metadata for multiple symbols in one call. payload format: {"symbol": str, "metadata": any}

• write_pickle

  Write arbitrary pickle-serializable data to a symbol.

• write_pickle_batch

  Write multiple pickle-serializable payloads in one call. payload format: {"symbol": str, "data": any, "metadata"?: any, "index_column"?: str}

• list_libraries

  List all libraries in the Arctic instance.

• create_library

  Create a new library with the given name. Libraries are the top-level containers for named symbols in ArcticDB.

• delete_library

  Delete a library and all its underlying data permanently. No-op if the library does not exist.

• library_exists

  Check whether a library with the given name exists.

• get_library

  Return the name and full list of symbols stored in a library.

• get_library_options

  Return non-enterprise library options.

• get_enterprise_options

  Return enterprise library options.

• reload_symbol_list

  Force a symbol-list cache reload for a library.

• compact_symbol_list

  Compact the symbol list cache into a single key in storage.

• prune_previous_versions

  Remove all non-snapshotted symbol versions except the latest.

• is_symbol_fragmented

  Check whether compaction would reduce segment count for a symbol.

• defragment_symbol_data

  Defragment symbol data by compacting fragmented row-sliced segments. Returns the resulting version information.

• query_filter

  Filter rows using conditions. filters: list of dicts, e.g. [{"column": "price", "op": ">", "value": 100}, {"column": "volume", "op": "<", "value": 5000}] Supported ops: >, >=, <, <=, ==, !=

• query_filter_isin

  Filter rows where column value is in a list. e.g. column="status", values=["active", "pending"]

• query_groupby

  Group by a column and aggregate. aggregations: e.g. {"price": "mean", "volume": "sum"} Supported agg ops: mean, sum, min, max, count

• query_date_range

  Filter rows by date range on a datetime-indexed symbol. start/end: ISO format strings e.g. "2024-01-01"

• query_resample

  Resample a datetime-indexed symbol and aggregate. rule: pandas offset string e.g. "1h", "1D", "1W" aggregations: e.g. {"price": "mean", "volume": "sum"} Supported agg ops: mean, sum, min, max, count, first, last

• list_symbols

  List all symbol names stored in the given library.

• read_symbol

  Read a symbol's full data as a list of row records, index included. Use as_of to read a specific historical version number; omit for the latest version.

• read_symbol_page

  Read a paginated slice of a symbol using backend row slicing and column projection. Returns rows along with the resolved version and pagination continuation metadata. Parameters: - library: Name of the library. - symbol: Name of the symbol. - offset: 0-indexed row offset to begin reading from (must

• head_symbol

  Return the first n rows of a symbol (default 5), index included. Useful for previewing large datasets without loading all data.

• tail_symbol

  Return the last n rows of a symbol (default 5), index included. Useful for previewing the most recent data without loading all data.

• write_symbol

  Write data to a symbol, creating a new version. data is a list of row records, where each record is a dict of column name to value. If a datetime column is detected (date, datetime, timestamp, ts, time, index), it is automatically set as the index to preserve time series structure. Example: [{"date"

• append_symbol

  Append rows to an existing symbol, creating a new version. The new rows must have index values strictly later than the existing data. data is a list of row records, where each record is a dict of column name to value.

• update_symbol

  Overwrite a date range of a timeseries symbol with new data, creating a new version. The symbol must have a datetime index. Only the rows whose index falls within the range of the provided data are replaced; rows outside that range are preserved. data is a list of row records, where each record is a

• delete_symbol

  Delete a symbol and all its versions from the library. No-op if the symbol does not exist.

• delete_data_in_range

  Delete all rows within a date range from a timeseries symbol, creating a new version. The symbol must have a datetime index. start and end are ISO 8601 date strings, e.g. "2024-01-01".

• symbol_exists

  Check whether a symbol exists in the given library.

• get_symbol_info

  Return descriptive information about a symbol: row count, column names, index type, date range, last update time, and metadata. Raises ValueError if the symbol does not exist.

• read_metadata

  Read only the metadata for a symbol without loading its data. Faster than a full read when only metadata is needed.

• write_metadata

  Update the metadata for a symbol without modifying its data. Creates a new version. The metadata must be a JSON-serialisable dict.

• list_versions

  List all versions of a symbol with their version number and write timestamp. Versions are created on every write, append, update, or metadata change.

• create_snapshot

  Create a named snapshot of the current library state. A snapshot captures the latest version of every symbol at the time of creation, allowing point-in-time reads across the whole library. Raises an error if a snapshot with that name already exists.

• list_snapshots

  List the names of all snapshots in a library.

• delete_snapshot

  Delete a named snapshot from a library. Only the snapshot reference is removed; the underlying symbol data is not affected. Raises an error if the snapshot does not exist.

• read_symbol_from_snapshot

  Read a symbol as it existed at the time a named snapshot was created. Returns the symbol's data as a list of row records.