xr_ai_tools.text_memory#

Participant-scoped persistent text-memory tools.

Classes#

`AddTranscriptRequest`	One timestamped transcript segment to persist.
`AddTranscriptResult`	Confirmation that a transcript segment was persisted.
`TranscriptSegment`	Persisted transcript text at a point in time.
`QueryTranscriptsRequest`	Source and inclusive time window for a transcript query.
`QueryTranscriptsResult`	Transcript segments selected from one source and time window.
`ListTranscriptSourcesResult`	Persistent source identifiers known to text memory.
`TranscriptStatsRequest`	Select one transcript source for statistics.
`TranscriptStatsResult`	Aggregate storage and time-range statistics for one source.
`RecallConversationRequest`	Participant and inclusive time window for conversation recall.
`ConversationEntry`	One timestamped user or agent turn in a recalled conversation.
`RecallConversationResult`	Chronological user and agent turns recalled for a participant.
`TextMemoryTool`	Append timestamped text to participant-scoped JSONL storage.
`TextMemoryTools`	Own transcript persistence, query, statistics, and conversation recall tools.

Module Contents#

class xr_ai_tools.text_memory.AddTranscriptRequest#

One timestamped transcript segment to persist.

source_id: str#: Participant or internal source identifier.

timestamp_us: int#: Unix timestamp in microseconds.

text: str#: Text segment to persist.

classmethod text_must_not_be_blank(value: str) → str#: Require visible text in a persisted segment.

class xr_ai_tools.text_memory.AddTranscriptResult#

Confirmation that a transcript segment was persisted.

ok: bool = True#: Whether the append operation succeeded.

class xr_ai_tools.text_memory.TranscriptSegment#

Persisted transcript text at a point in time.

timestamp_us: int#: Unix timestamp in microseconds.

text: str#: Persisted transcript text.

class xr_ai_tools.text_memory.QueryTranscriptsRequest#

Source and inclusive time window for a transcript query.

source_id: str#: Participant or internal source identifier.

start_us: int#: Inclusive window start in Unix microseconds.

end_us: int#: Inclusive window end in Unix microseconds.

class xr_ai_tools.text_memory.QueryTranscriptsResult#

Transcript segments selected from one source and time window.

segments: list[TranscriptSegment]#: Matching segments ordered by timestamp.

class xr_ai_tools.text_memory.ListTranscriptSourcesResult#

Persistent source identifiers known to text memory.

sources: list[str]#: Source identifiers in lexical order.

class xr_ai_tools.text_memory.TranscriptStatsRequest#

Select one transcript source for statistics.

source_id: str#: Participant or internal source identifier.

class xr_ai_tools.text_memory.TranscriptStatsResult#

Aggregate storage and time-range statistics for one source.

source_id: str#: Participant or internal source identifier.

count: int#: Number of stored transcript segments.

total_chars: int#: Total number of characters across stored segments.

earliest_us: int | None#: Earliest stored Unix timestamp in microseconds, if present.

latest_us: int | None#: Latest stored Unix timestamp in microseconds, if present.

class xr_ai_tools.text_memory.RecallConversationRequest#

Participant and inclusive time window for conversation recall.

participant_id: str#: Participant whose conversation to recall.

start_us: int#: Inclusive window start in Unix microseconds.

end_us: int#: Inclusive window end in Unix microseconds.

class xr_ai_tools.text_memory.ConversationEntry#

One timestamped user or agent turn in a recalled conversation.

timestamp_us: int#: Unix timestamp in microseconds.

role: Literal['user', 'agent']#: Speaker responsible for the text.

text: str#: Conversation text.

class xr_ai_tools.text_memory.RecallConversationResult#

Chronological user and agent turns recalled for a participant.

entries: list[ConversationEntry]#: Conversation turns ordered by timestamp.

class xr_ai_tools.text_memory.TextMemoryTool(directory: str | pathlib.Path | _TranscriptStore)#: Append timestamped text to participant-scoped JSONL storage.

class xr_ai_tools.text_memory.TextMemoryTools(directory: str | pathlib.Path)#: Own transcript persistence, query, statistics, and conversation recall tools.