Models API

This page documents the supported data-model boundary used by integrations. The CLI remains more stable than direct model construction. Model operations are side-effect free unless a caller passes them to storage or workflow functions.

.. py:module:: memoryledger.models

.. py:class:: EvidenceRef(kind, title, uri, excerpt=’’, content_hash=’’, line_start=None, line_end=None, timestamp=’’)

module:

memoryledger.models

Bases: :py:class:object

A reviewable pointer to source evidence for a memory.

type kind:

sphinx_autodoc_typehints_type:

\:py\:class\:\str``

param kind:

type title:

:sphinx_autodoc_typehints_type:\:py\:class\:\str``

param title:

type uri:

:sphinx_autodoc_typehints_type:\:py\:class\:\str``

param uri:

type excerpt:

:sphinx_autodoc_typehints_type:\:py\:class\:\str``

param excerpt:

type content_hash:

:sphinx_autodoc_typehints_type:\:py\:class\:\str``

param content_hash:

Import example:

from memoryledger.models import Memory, RenderConfig

Validation errors are raised by guardrails and storage layers; model dataclasses do not perform workspace writes or path resolution.