memleaf 0.2.51__tar.gz → 0.2.53__tar.gz

{memleaf-0.2.51 → memleaf-0.2.53}/CHANGELOG.md

@@ -2,6 +2,21 @@
 
 All notable changes to memleaf are documented here.
 
+## 0.2.53 — 2026-09-13
+
+- Keep the memory when only the project name is unproven. A real turn was lost in full because of one word: the user wrote "在弄个记账的小玩意儿，就扔家里那台 N100 上跑，没打算上云", the model named the project `project:记账小玩意儿` -- the user's phrase with 的 removed -- and Core refuses a project scope whose name is not in the candidate's own evidence. That refusal discarded the whole candidate, so the N100 deployment and the no-cloud decision went with the unproven name, leaving three evidence units unresolved and nothing written. The memory was grounded; only the affiliation was not. The claim is now dropped and the memory kept with `global`, which asserts no ownership and loses no source-backed content. The refusal itself is unchanged: Core still never records a project it cannot ground.
+- Name the project from the evidence verbatim, and prefer the shortest term that appears. The same turn showed the model composing a name by editing the user's words, which is the failure mode the grounding rule catches. The contract now says so directly, with the worked example: the thing described as "记账的小玩意儿" is `project:记账`, not `project:记账小玩意儿` and not `project:记账的小玩意儿`. Measured against the live model on the exact captured turn, the composed name is gone and both the with-assistant and user-only variants yield `project:记账`.
+- Make the B3 outcome counters actually surface. `b3_normalization_count` and `b3_candidate_deferred_count` were emitted but were not in the metric allowlist, so they were discarded without a word -- which made the "visible rather than silent" claim in the 0.2.49 and 0.2.50 notes untrue. Both are now recorded, together with `b3_ungrounded_scope_dropped_count` for the fallback above.
+
+Verification for this release used the real model on the captured turn. 118 tests pass with no model call, pinning the grounding rule that caused the loss (a verbatim name passes, the edited name does not), the fallback (a project scope becomes `global`, non-project scopes survive, an empty input still yields a legal scope), and the grounding decision itself. Against the live route, replaying the exact captured turn with and without the assistant's reply now yields `project:记账` in both cases, where before the fix the with-assistant variant yielded the ungrounded composed name. The deferred candidates of that already-processed turn are replayed from the journal rather than re-attempted, so the memory lost by the old code is not recovered by upgrading; the fix applies to turns processed from now on.
+
+## 0.2.52 — 2026-09-13
+
+- Stop an assistant message from originating a durable claim. A three-sentence acceptance run recorded a project requirement the user never asked for: the user wrote only "我写日志习惯用中文，英文的扫两眼就跳过去了", the assistant replied with its own elaboration ("记账工具的日志……都写成中文；异常堆栈上头再包一层人话"), and the planner stored that elaboration as a project fact. The design allows an assistant reply as a source for *organising what the user already said*, so the rule is now explicit about what that excludes: an assistant message may organise, restate or report what the user's own evidence establishes, may never be the only support for a durable claim, and contributes nothing where it states a decision, requirement, rule, plan, recommendation or added detail of its own -- however definite it sounds. Consequences the assistant derived from a user preference are the assistant's, not the user's, and are not recorded as requirements or project state.
+- Fix a regression the scope vocabulary introduced, before it could reach a project-scoped Vault. The scoped fallback runs when a turn matches no existing memory and the session scope is specific, and it declared itself ambiguous when the scope held more than one record -- ``len(records) > 1``. A populated scope therefore always reported ambiguity, which cleared ``lookup_complete`` and deferred every CREATE in that turn. It stayed dormant while sessions were ``global``-only, because ``_has_specific_scope`` is false for ``global`` and the fallback never ran; the first session that carried ``project:记账`` activated it, and turn three of the acceptance run deferred with ``lookup_incomplete`` and wrote nothing with zero rows withheld from the catalog. Ambiguity now means what it should: this turn names more than one specific scope, so the fallback cannot tell which one owns it.
+
+Verification for this release used the real model and a real Vault. 110 tests pass with no model call, covering the assistant-source rule and the allowance it must preserve, and the scope fallback across a populated scope, two named projects, global-only and an empty scope. Against the live route, the assistant-source rule was checked on three cases: the reproduced defect (the user's preference is now recorded alone, the assistant's elaboration is marked ``assistant_restatement`` and nothing is written from it), an assistant reply that organises user-stated progress, and a requirement the user explicitly accepts -- the two controls still work. The ambiguity fix was checked against the deferred production turn, which now reports ``lookup_complete: true`` with all 21 in-scope records projected. The assistant-source rule is published in the prompt and enforced by the planner's judgement, not by a Core-side structural check, so a model that ignores it can still record such a claim; no measurement of that rate was made.
+
 ## 0.2.51 — 2026-09-13
 
 - Publish the scope vocabulary in the B3 single-pass contract. The compact contract described `scopes` only as `nonempty string[]` and never named `project:<name>`, so a planner handed an empty `scope_registry` had exactly one value it could be sure was legal. A first message that plainly establishes a project -- "最近在弄个记账的小玩意儿，就扔家里那台 N100 上跑" -- was filed under `global`, producing the internally inconsistent `type=project` beside `scopes=["global"]`. The legacy summarize contract always carried this paragraph; B3 lost it, the same way it had lost the todo-field and date rules repaired in 0.2.48 and 0.2.50.

{memleaf-0.2.51/src/memleaf.egg-info → memleaf-0.2.53}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: memleaf
-Version: 0.2.51
+Version: 0.2.53
 Summary: A local-first Markdown memory core for AI agents
 Author: memleaf contributors
 License-Expression: MIT
@@ -23,7 +23,7 @@ Dynamic: license-file
 
 [English](README.en.md) · [PyPI](https://pypi.org/project/memleaf/) · [GitHub](https://github.com/miffyblueboo/memleaf)
 
-> **版本：0.2.51。**
+> **版本：0.2.53。**
 > 记忆只从用户与 Agent 的可见对话提炼。Agent 已在回复中整理的事实、项目进展和明确待办可作为来源；邮件、附件、网页和终端等工具原文不进入记忆提炼。模型负责保留原话中的不确定性，现有记忆仅用于比较、去重和更新。Markdown 仍是唯一事实源。
 > **当前版本支持 Hermes 和 Codex。** Antigravity（反重力）不检测、不安装、不配置。
 

{memleaf-0.2.51 → memleaf-0.2.53}/README.en.md

@@ -4,7 +4,7 @@
 
 [中文](README.md) · [PyPI](https://pypi.org/project/memleaf/) · [GitHub](https://github.com/miffyblueboo/memleaf)
 
-> **Version: 0.2.51.**
+> **Version: 0.2.53.**
 > Automatic extraction now uses only the current turn's visible user input and final assistant reply. Raw tool output, attachments, web/file/terminal payloads and legacy tool-evidence bodies are not new source evidence; existing or retrieved memory remains comparison context rather than source authority. Background processing is persisted as a local job and can be checked through the read-only `process_status` MCP tool; failed work remains retryable and fail closed. The release also tightens source/date grounding, target reconciliation, duplicate/no-op handling and semantic review before writes. Markdown remains the sole source of truth with no SQLite runtime dependency. Acceptance covers deterministic regression suites and synthetic inputs; it does not claim real-mail or customer-business acceptance.
 > **The current release supports Hermes and Codex.** Antigravity is not detected, installed, or configured.
 

{memleaf-0.2.51 → memleaf-0.2.53}/README.md

@@ -4,7 +4,7 @@
 
 [English](README.en.md) · [PyPI](https://pypi.org/project/memleaf/) · [GitHub](https://github.com/miffyblueboo/memleaf)
 
-> **版本：0.2.51。**
+> **版本：0.2.53。**
 > 记忆只从用户与 Agent 的可见对话提炼。Agent 已在回复中整理的事实、项目进展和明确待办可作为来源；邮件、附件、网页和终端等工具原文不进入记忆提炼。模型负责保留原话中的不确定性，现有记忆仅用于比较、去重和更新。Markdown 仍是唯一事实源。
 > **当前版本支持 Hermes 和 Codex。** Antigravity（反重力）不检测、不安装、不配置。
 

{memleaf-0.2.51 → memleaf-0.2.53}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "memleaf"
-version = "0.2.51"
+version = "0.2.53"
 description = "A local-first Markdown memory core for AI agents"
 readme = "README.md"
 requires-python = ">=3.11"

{memleaf-0.2.51 → memleaf-0.2.53}/src/memleaf/__init__.py

@@ -1,6 +1,6 @@
 """Local-first Markdown memory core for AI agents."""
 
-__version__ = "0.2.51"
+__version__ = "0.2.53"
 
 from .config import DEFAULT_CONFIG, default_config, load_config, save_config
 from .frontmatter import FrontmatterError, dump_frontmatter, dump_yaml, load_yaml, parse_frontmatter

{memleaf-0.2.51 → memleaf-0.2.53}/src/memleaf/hermes_provider/plugin.yaml

@@ -1,5 +1,5 @@
 name: memleaf
-version: 0.2.51
+version: 0.2.53
 description: "Hermes-native external MemoryProvider for local-first Markdown memory shared with the memleaf MCP server."
 hooks:
   - prefetch

{memleaf-0.2.51 → memleaf-0.2.53}/src/memleaf/model_execution.py

@@ -51,6 +51,12 @@ _COVERAGE_DIAGNOSTIC_FIELDS = frozenset({"explanation"})
 _METRIC_EVENT_FIELDS = frozenset({
     "repair_attempted_count", "repair_rejected_semantic_drift_count",
     "parse_accepted_count", "decision_case_normalization_count",
+    # Normalization, deferral and scope fallback are all outcomes a turn can
+    # reach without failing, so they must actually surface in the metrics.  The
+    # first two were emitted but silently dropped here, which made the
+    # "visible rather than silent" claim in their release notes untrue.
+    "b3_normalization_count", "b3_candidate_deferred_count",
+    "b3_ungrounded_scope_dropped_count",
 })
 _THINKING_OBSERVATION_SOURCES = frozenset({
     "request_parameter", "reasoning_tokens", "reasoning_content", "unavailable",
@@ -82,6 +88,9 @@ def _metric_bucket() -> dict[str, Any]:
         "repair_rejected_semantic_drift_count": 0,
         "parse_accepted_count": 0,
         "decision_case_normalization_count": 0,
+        "b3_normalization_count": 0,
+        "b3_candidate_deferred_count": 0,
+        "b3_ungrounded_scope_dropped_count": 0,
         "_first_started": None,
         "_last_finished": None,
     }

{memleaf-0.2.51 → memleaf-0.2.53}/src/memleaf/planning_context.py

@@ -393,7 +393,21 @@ class PlanningContext:
 
 
     def _scope_records_unlocked(self, scope: Any) -> tuple[list[Any], bool]:
-        """Read and rank active records once for a scoped fallback."""
+        """Read and rank active records once for a scoped fallback.
+
+        ``ambiguous`` answers a question about *scope identity*: does this turn
+        name more than one specific scope, so that the fallback cannot tell
+        which one owns it?  It deliberately does not count how many memories the
+        scope holds.  A scope carrying many memories is ordinary, and a session
+        naming one project owns its turns unambiguously.
+
+        Counting records instead made the fallback declare itself ambiguous
+        whenever a scope was populated, which marked the lookup unprovable and
+        deferred every CREATE in a project-scoped session -- while a
+        ``global``-only session, where the fallback never runs, was unaffected.
+        The catalog ceiling is what bounds how much context is safe to supply;
+        ambiguity is not.
+        """
 
         active_records = self.service._read_memories_unlocked("knowledge")
         scoped = filter_by_scope(
@@ -415,7 +429,13 @@ class PlanningContext:
             ),
             reverse=True,
         )
-        return records, len(records) > 1
+        requested = [scope] if isinstance(scope, str) else list(scope or [])
+        specific = {
+            value
+            for value in requested
+            if isinstance(value, str) and value not in {"", "global", "unscoped"}
+        }
+        return records, len(specific) > 1
 
 
     @classmethod

{memleaf-0.2.51 → memleaf-0.2.53}/src/memleaf/single_pass_memory_planner.py

@@ -34,6 +34,25 @@ from .validation import ModelOutputError, parse_summarize_output
 from .llm import ModelUnavailable
 
 
+def _drop_ungrounded_project_scopes(scopes: Iterable[Any]) -> list[str]:
+    """Return the scopes that assert no unproven project ownership.
+
+    A project scope is a claimed affiliation and Core refuses one the evidence
+    does not name.  The memory carrying it is usually grounded in full; only the
+    affiliation is unproven, so the claim is dropped rather than the memory.
+    Measured on a real turn, discarding the candidate instead lost
+    "在弄个记账的小玩意儿，跑在 N100，不上云" entirely because the model had
+    named the project "记账小玩意儿" while the user wrote "记账的小玩意儿".
+    """
+
+    kept = [
+        scope
+        for scope in scopes
+        if not (isinstance(scope, str) and scope.startswith("project:"))
+    ]
+    return kept or ["global"]
+
+
 class SinglePassMemoryPlanner(MemoryPlanner):
     """One semantic model call for an ordinary automatic turn."""
 
@@ -430,10 +449,24 @@ class SinglePassMemoryPlanner(MemoryPlanner):
                 validation_scope_registry,
                 authorized_project_scopes,
             ):
-                raise ModelOutputError(
-                    "B3 project scope is not grounded by claimed source",
-                    validation_detail="scope_not_grounded",
-                )
+                # The memory itself is grounded; only the ownership claim is
+                # not.  Discarding the whole candidate over an unproven project
+                # name loses source-backed content the user stated, so the
+                # claim is dropped and the memory is kept unscoped instead.
+                # Nothing is fabricated either way: no ownership is asserted.
+                kept = _drop_ungrounded_project_scopes(scopes)
+                dropped = kept != list(scopes)
+                scopes = kept
+                scope_source = self._derived_scope_source(scopes, scope_background, scope)
+                candidate["scopes"] = scopes
+                candidate["scope_source"] = scope_source
+                if dropped:
+                    event = getattr(self.model, "_record_metric_event", None)
+                    if callable(event):
+                        event(
+                            {"stage": "single_pass", "operation": "single_pass_primary"},
+                            "b3_ungrounded_scope_dropped_count",
+                        )
             if (
                 decision == "UPDATE"
                 and target_memory is not None

{memleaf-0.2.51 → memleaf-0.2.53}/src/memleaf/single_pass_plan.py

@@ -121,6 +121,7 @@ SINGLE_PASS_SYSTEM = f"""You are memleaf's single-pass memory planner.
 
 SOURCE
 Only current_evidence may establish new facts or changes. local_memory_catalog and native_memory_catalog are comparison context; native memory is never an UPDATE/NO_CHANGE target. Never invent source facts, ownership, dates, status, obligations, numbers, relationships or IDs.
+An assistant message is evidence of what you reported, never of what the user wants. It may organise, restate or report what the user's own evidence establishes, but it may never be the only support for a durable claim. When it states something the user's message does not establish -- your own decision, requirement, rule, plan, recommendation or added detail -- record nothing from that statement, however definite it sounds. Consequences you derived from a user preference are yours, not the user's: never record them as a requirement or as project state. Only something the user stated or accepted becomes memory.
 
 TASK
 Extract every independently useful long-term memory. Keep independently retrievable/updateable topics separate and preserve entity, condition, polarity, uncertainty, ownership, state and meaning-critical numbers/codes. CREATE only when no supplied local memory represents the durable information; UPDATE only when current evidence proves a change to one supplied local memory; NO_CHANGE only when it adds no semantic change; DEFERRED for a durable candidate that cannot safely reach a terminal decision. Project ownership and platform/system names are separate judgments. Do not turn every negation into no_memory and do not use NO_CHANGE to hide ambiguity.
@@ -128,7 +129,7 @@ Extract every independently useful long-term memory. Keep independently retrieva
 SCOPES
 Legal values: global | domain:<name> | portfolio:<name> | project:<name> | unscoped. scopes is a nonempty array; at most one project:<name> per memory; unscoped must be the only value, and Core then records insufficient_context.
 When a candidate concerns one distinct project the user is working on -- a type=project memory, or the work, decisions, state or deadlines that belong to that project -- scope it to that project as project:<name>, using the name the evidence itself uses. The project does not need to be registered first; scope_registry only lists what already exists.
-Ground the name in this candidate's own cited evidence, or in a project scope explicitly supplied for this turn. A product, platform, system, vendor, notification source, comparison or implementation context is not ownership by name alone.
+Ground the name in this candidate's own cited evidence, or in a project scope explicitly supplied for this turn. Never edit, translate or combine the evidence's words into a name that does not appear there, and choose the shortest term that does appear and names the project: the thing described as "记账的小玩意儿" is project:记账, not project:记账小玩意儿 and not project:记账的小玩意儿. An ungrounded name is rejected and its memory is deferred instead of written. A product, platform, system, vendor, notification source, comparison or implementation context is not ownership by name alone.
 Use global only for a fact that no single project owns: a standing personal preference, a machine-wide or tool-wide rule, or an environment fact. Never invent, translate or borrow a project name.
 
 DATES

{memleaf-0.2.51 → memleaf-0.2.53/src/memleaf.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: memleaf
-Version: 0.2.51
+Version: 0.2.53
 Summary: A local-first Markdown memory core for AI agents
 Author: memleaf contributors
 License-Expression: MIT
@@ -23,7 +23,7 @@ Dynamic: license-file
 
 [English](README.en.md) · [PyPI](https://pypi.org/project/memleaf/) · [GitHub](https://github.com/miffyblueboo/memleaf)
 
-> **版本：0.2.51。**
+> **版本：0.2.53。**
 > 记忆只从用户与 Agent 的可见对话提炼。Agent 已在回复中整理的事实、项目进展和明确待办可作为来源；邮件、附件、网页和终端等工具原文不进入记忆提炼。模型负责保留原话中的不确定性，现有记忆仅用于比较、去重和更新。Markdown 仍是唯一事实源。
 > **当前版本支持 Hermes 和 Codex。** Antigravity（反重力）不检测、不安装、不配置。