PowercontextBuiltinArtifactsmemoryprotocols

MemoryBackend

Functions

funcany_tagged_entry_ids(memory) -> frozenset[str]

Return logical entry IDs carrying any tag in this Memory.

parammemoryArtifactRef

Returns

frozenset[str]
functagged_entry_ids(memory, tag_filter) -> frozenset[str]

Return exact tag matches without imposing a candidate limit.

parammemoryArtifactRef
paramtag_filterTagFilter

Returns

frozenset[str]
funccapabilities() -> MemoryCapabilities

Return probed deployment capabilities.

Returns

powercontext.builtin.artifacts.memory.models.MemoryCapabilities
funcget(memory) -> Memory

Load one exact Memory Revision.

parammemoryArtifactRef

Returns

powercontext.builtin.artifacts.memory.models.Memory
funclatest(artifact_id) -> Memory

Load the current head of one Memory identity.

paramartifact_idstr

Returns

powercontext.builtin.artifacts.memory.models.Memory
funcentries(memory) -> tuple[MemoryEntryVersion, ...]

Load the entry versions referenced by an exact manifest.

parammemoryArtifactRef

Returns

tuple[powercontext.builtin.artifacts.memory.models.MemoryEntryVersion, ...]
funcprojections(memory) -> tuple[MemoryProjection, ...]

Load rebuildable active-head projections for an exact Memory Revision.

parammemoryArtifactRef

Returns

tuple[powercontext.builtin.artifacts.memory.protocols.MemoryProjection, ...]
funcrebuild_projections(embedding_model=None) -> None

Rebuild active-head and search projections from authoritative revisions.

paramembedding_modelEmbeddingModel | None
= None

Returns

None
funcbegin() -> AbstractAsyncContextManager[MemoryUnitOfWork]

Open the adapter-specific atomic write boundary.

Returns

contextlib.AbstractAsyncContextManager[powercontext.builtin.artifacts.memory.protocols.MemoryUnitOfWork]
funcchanges(memory, since_revision) -> tuple[MemoryRevisionChanges, ...]

Read compact Revision changes without entry bodies.

parammemoryArtifactRef
paramsince_revisionint | None

Returns

tuple[powercontext.builtin.artifacts.memory.models.MemoryRevisionChanges, ...]
funcvector_complete(memories, profile) -> bool

Derive fixed-profile vector completeness for selected heads.

parammemoriestuple[ArtifactRef, ...]
paramprofileEmbeddingProfile

Returns

bool
funcsearch(request) -> MemorySearchChannels

Return backend-ordered FTS/vector channels after manifest checks.

paramrequestMemorySearchRequest

Returns

powercontext.builtin.artifacts.memory.protocols.MemorySearchChannels
funcexpand(hits) -> tuple[MemoryEntryVersion, ...]

Load and validate exact versions anchored by hits.

paramhitstuple[MemoryHit, ...]

Returns

tuple[powercontext.builtin.artifacts.memory.models.MemoryEntryVersion, ...]

CandidatePipeline class

Previous Page

MemoryCandidateRequest class

Next Page

On this page

Functions