Episodic memory: boundary-based segmentation, facts/episode retrieval separation, and v2 extraction default

.env.template

@@ -29,7 +29,7 @@ USER_SUMMARY_EVERY_N=20
 # the SDK and the Function App are deployed against the same database.
 #
 #   * Unset / blank        -> SDK auto-trigger fires; FA change-feed SKIPS.
-#                             (Pure SDK deployments — no env config needed.)
+#                             (Pure SDK deployments - no env config needed.)
 #   * "inprocess"          -> SDK auto-trigger fires; FA change-feed SKIPS.
 #   * "durable"            -> SDK auto-trigger SKIPS; FA change-feed fires.
 #

.github/workflows/_test_release.yml

@@ -85,7 +85,7 @@ jobs:
           verbose: true
           print-hash: true
           repository-url: https://test.pypi.org/legacy/
-          # CI-only — overwrites a same-version file if a re-run is needed.
+          # CI-only - overwrites a same-version file if a re-run is needed.
           # https://github.com/pypa/gh-action-pypi-publish#tolerating-release-package-file-duplicates
           skip-existing: true
           # Attestations default-on in v1.11.0+ and require additional

.github/workflows/release.yml

@@ -105,7 +105,7 @@ jobs:
         env:
           PKG_NAME: ${{ needs.build.outputs.pkg-name }}
           VERSION: ${{ needs.build.outputs.version }}
-        # Two-pass retry — TestPyPI mirror propagation can lag by a few seconds.
+        # Two-pass retry - TestPyPI mirror propagation can lag by a few seconds.
         # Primary index is real PyPI so all the dependencies resolve normally;
         # the just-published preview only lives on TestPyPI.
         run: |
@@ -164,7 +164,7 @@ jobs:
           packages-dir: dist/
           verbose: true
           print-hash: true
-          # Attestations default-on in v1.11.0+ — opt out until configured.
+          # Attestations default-on in v1.11.0+ - opt out until configured.
           attestations: false
 
   mark-release:

CHANGELOG.md

@@ -70,7 +70,7 @@
 * `ProcessThreadResult` gains `procedural` and `user_summary` fields. `extract_memories` returns a `dropped_episodic_count` for monitoring LLM-extraction quality.See [PR:#20](https://github.com/aayush3011/AgentMemoryToolkit/pull/20)
 
 
-## [0.1.0b1] — 2026-06-01
+## [0.1.0b1] - 2026-06-01
 
 
 Initial public preview release.
@@ -91,7 +91,7 @@ Pin a specific version when integrating.
   tag filters, and per-type scoping.
 - Built-in memory processing pipeline: fact extraction, thread/user
   summarization, procedural-memory synthesis, contradiction handling, and
-  deduplication — all driven by versioned `.prompty` prompts.
+  deduplication - all driven by versioned `.prompty` prompts.
 - Two processor backends: `InProcessProcessor` (default, runs in your
   application process) and `DurableFunctionProcessor` (offloads work to a
   sibling Azure Function app via Cosmos DB change feed).

Docs/RELEASING.md

@@ -19,7 +19,7 @@ This project uses [PEP 440](https://peps.python.org/pep-0440/) versioning:
 
 Before cutting a release:
 
-1. **CI green on `main`** — every workflow in `.github/workflows/` passes.
+1. **CI green on `main`** - every workflow in `.github/workflows/` passes.
 2. **Unit tests pass locally** in a fresh venv:
    ```bash
    python -m venv /tmp/release-venv
@@ -32,21 +32,21 @@ Before cutting a release:
    ```bash
    AGENT_MEMORY_RUN_INTEGRATION=true pytest tests/integration -q
    ```
-4. **Samples and notebooks work** — every script under `Samples/` runs to
+4. **Samples and notebooks work** - every script under `Samples/` runs to
    completion against the live environment.
-5. **No uncommitted local changes** — `git status` clean.
+5. **No uncommitted local changes** - `git status` clean.
 
 ## Cutting a release
 
 1. **Bump the version** in `pyproject.toml`:
-2. **Update `CHANGELOG.md`** — add a new section with the version, the
+2. **Update `CHANGELOG.md`** - add a new section with the version, the
    date, and a summary of changes. Move entries from the unreleased
    section if you keep one.
 3. **Bump the Function app's SDK pin** in `function_app/requirements.txt`
-   to match — `azure-cosmos-agent-memory==<NEW_VERSION>`. The FA installs
+   to match - `azure-cosmos-agent-memory==<NEW_VERSION>`. The FA installs
    the SDK from PyPI, so the pin must move in lockstep with the SDK
    release. (If the release workflow fails after merge, the FA will be
-   pinned to a non-existent version until you cut a follow-up patch —
+   pinned to a non-existent version until you cut a follow-up patch -
    coordinate the merge + release-workflow run together.)
 4. **Open a PR** with the version bump + CHANGELOG + updated FA pin.
    Suggested title: `Release v<NEW_VERSION>`. Get it reviewed and merged
@@ -55,7 +55,7 @@ Before cutting a release:
    - Navigate to **Actions → release → Run workflow**
    - Pick the `main` branch
    - Click **Run workflow**
-6. **The workflow does the rest** — see `.github/workflows/release.yml`:
+6. **The workflow does the rest** - see `.github/workflows/release.yml`:
    - Builds `dist/*.whl` + `dist/*.tar.gz`
    - Asserts no namespace `__init__.py` shadows are in the wheel
    - Publishes to **TestPyPI** via trusted publishing
@@ -69,7 +69,7 @@ Before cutting a release:
 ## Namespace package note
 
 `azure-cosmos-agent-memory` installs files under `azure/cosmos/agent_memory/`.
-It MUST NOT ship `azure/__init__.py` or `azure/cosmos/__init__.py` — those
+It MUST NOT ship `azure/__init__.py` or `azure/cosmos/__init__.py` - those
 are owned by the `azure-cosmos` package. The wheel build is configured
 (`[tool.setuptools.packages.find]` with `include = ["azure.cosmos.agent_memory*"]`
 and `namespaces = true`) so that only the `agent_memory` subtree is packaged.
@@ -94,4 +94,4 @@ If a release ships a regression:
 3. Add a deprecation note to the GitHub Release pointing at the
    replacement version.
 
-Never delete a PyPI release — yank it and ship a fixed version instead.
+Never delete a PyPI release - yank it and ship a fixed version instead.

Docs/azure_testing.md

@@ -147,7 +147,7 @@ Set any threshold to `"0"` to disable that processing type.
 
 The `leases` container is provisioned by `create_memory_store()` alongside the `memories` and `counter` containers, so the Function App should be configured to use that existing lease container.
 
-The Function App authenticates to Cosmos DB and Azure OpenAI via its managed identity — there's no shared key or function-key handoff between the SDK and the Function App.
+The Function App authenticates to Cosmos DB and Azure OpenAI via its managed identity - there's no shared key or function-key handoff between the SDK and the Function App.
 
 ---
 
@@ -300,7 +300,7 @@ print(memory.search_cosmos("hello", user_id="user-1"))
 
 ### Durable processing (change-feed driven)
 
-Processing is no longer invoked directly from the SDK — write turns with `add_cosmos()` / `push_to_cosmos()` and the deployed Function App's change-feed trigger fires the `extract_memories`, `thread_summary`, and `user_summary` orchestrators per the configured thresholds.
+Processing is no longer invoked directly from the SDK - write turns with `add_cosmos()` / `push_to_cosmos()` and the deployed Function App's change-feed trigger fires the `extract_memories`, `thread_summary`, and `user_summary` orchestrators per the configured thresholds.
 
 ```python
 # Write enough turns to cross THREAD_SUMMARY_EVERY_N (default 10).