MemoryGuard

Governed shared memory for coding agents.
Local-first MCP memory with automatic organization, scoped rules, evidence, and rollback.

中文文档

Let agents write without turning shared memory into an unreviewed pile. MemoryGuard organizes each write, preserves the evidence behind changes, and keeps governance decisions reversible.

No account. No remote server. No telemetry. Memory stays local.

Quick start · Upgrade · Knowledge Library · Architecture · Supported hosts · Privacy and safety

_{A synthetic governed projection: signals move through memory categories while raw conversation text remains outside the graph.}

What's New in v0.7.7

v0.7.7 is a focused provider-repair and runtime-alignment hotfix:

• Safe bare-shell repair: memoryguard provider repair ... can use an in-process trusted repair capability when the canonical control home is verified and exactly one active binding is present. Missing or ambiguous context remains fail-closed, and other CLI mutations are unchanged.
• One interpreter for installed repairs: non-editable wheel repairs ignore inherited stale snapshots and write the current sys.executable to the Codex MCP entry and all seven lifecycle Hooks. Editable/local installs keep their content-keyed immutable snapshot behavior.
• Binding continuity: repairs preserve the existing Agent identity and shared-group binding.

See the v0.7.7 release record.

What's New in v0.7.6

v0.7.6 is a focused Codex Hook/MCP runtime reliability release:

• One runtime snapshot: Codex lifecycle Hooks and MCP now share the same content-keyed, immutable runtime snapshot, avoiding PATH/package-metadata version drift.
• Concurrent tool safety: session-state writes use short, re-entrant cross-process lock windows instead of locking around the whole Hook run, so concurrent tools and sub-agent dispatch do not time out behind one another.
• Consistent bootstrap state: a failed bootstrap clears stale success state; recovery clears old errors and records a stable context hash. Ordinary Stop failures stay fail-open, while explicit mandatory-budget overflow is retained for the next tool call to fail closed.
• Stable Agent identity: the built-in Grok profile is discovered from its installed user surface. Verified provider/CLI/MCP metadata maps to a stable program identity; generic or opaque values remain unresolved instead of being guessed.
• Clearer governance shell: the GUI follows the seven-page reference information architecture with persistent navigation, a central workspace, and a context rail. Agent cards show readable names, source summaries, and safe family-aware icon fallbacks; technical IDs remain in collapsed details.

See the v0.7.6 release record.

What's New in v0.7.5

v0.7.5 is a focused conflict-review correctness release:

• Readable conflict snapshots: conflict members include bounded, redacted previews from the V2 atom/tombstone history, so the review view remains useful without exposing raw bodies.
• Clear reasons and safe candidates: known conflict reasons have Chinese labels, while deleted, superseded, rejected, quarantined, shadowed, missing, or otherwise invalid members cannot be selected as the keeper.
• Fail-closed resolution: a conflict can be resolved only when at least two members are still live and selectable; stale historical groups remain visible for audit instead of becoming mutation targets.
• Honest counts: historical conflict-group totals stay separate from the actionable count used by the governance overview.

See the v0.7.5 release record.

What's New in v0.7.4

v0.7.4 is a governance correctness and runtime reliability release:

• Canonical updates: related rules, habits, and memories merge into one evolving canonical record while graph branches, provenance, scope, evidence, and reversible history remain available.
• Bounded mandatory context: more than 20 mandatory rules raises a health warning instead of truncating storage or injection; character/token budget, sensitive-content, and corrupt-state failures remain fail-closed.
• Readable governance: Agent identities, automatic decisions, and risk signals use readable labels and explanations. GUI governance groups start collapsed; conflict and risk summaries open their details; risk entries explain reason, impact, and action; CodeGraph shares the same governed identity/scope/evidence path.
• Reliable audit state: run_audit and get_audit return explicit completion state and timezone-aware timestamps. The GUI no longer leaves a completed scan at 待扫描 and keeps the reader popover above navigation; missing numeric health evidence remains unknown instead of becoming a fabricated score. Governance phases show readable completed/current/ pending/undetermined states.
• Codex runtime safety: Hook/bootstrap errors are classified accurately; NO_SOURCE is a neutral fallback. Provider installation uses content-keyed, non-editable MCP snapshots with atomic failure behavior, packaged static asset refresh, and PEP 610 copied-install diagnostics.
• Safe mutation and migration: GUI partial mutations and evidence conflicts preserve the last valid state. System group_outbox projection advances event and checkpoint atomically; safe repair handles old lag while pending/failed events remain fail-closed. Migration replay contracts verify idempotent retries and failure evidence.

Validation for this release uses the existing incremental CI tail (447/447 passed), related focused suites, and git diff --check; the full suite was not rerun for this release preparation.

See the v0.7.4 release record.

What's New in v0.7.3

v0.7.3 is a focused shared-history visibility fix.

See the v0.7.3 release record.

What's New in v0.7.2

v0.7.2 is a focused correctness and compatibility release:

• Write → read-back and scope: trusted Agent/project scope now stays aligned across the write and read paths, so a successful write can be read back under the same scope without weakening isolation.
• Codex Hook lifecycle and transport: lifecycle cleanup uses verified Codex thread/workspace evidence, keeps ordinary Stop resumable, and reconciles terminal or deleted threads conservatively. The transport repair normalizes the current interpreter and UTF-8 stdio, then verifies initialize, tools/list, memory_status, and memory_read.
• Python 3.10 compatibility: the Codex transport repair path uses the project's memoryguard.toml_compat interface, selecting stdlib tomllib where available and tomli on Python 3.10.

See the v0.7.2 release record.

What's New in v0.7.1

v0.7.1 closes the migration and desktop lifecycle gaps found after the V2-only cutover:

• One-command migration: after upgrading the Python package, run memoryguard upgrade. The bare command uses the canonical user data home, migrates and verifies bindings, groups, memory, rules, history, and sources, activates V2, then removes only the successful migration backup batch. memoryguard upgrade --preview is the zero-write inspection path.
• One canonical control workspace: bare memoryguard gui, doctor, mcp-status, hooks, groups, and storage commands all use the same user-level data home, independent of the terminal's current directory.
• Recovered Agent and Group control: migrated bindings and groups remain editable; discovered but unbound Agents can enable a personal memory layer. Discovery exposes registered product/profile surfaces without guessing every directory on the machine or reading source bodies.
• Real projection engines: the build dialog lists only executable local Agent CLIs such as Cursor Agent and Codex. A selected CLI runs the governed extraction/enrichment pipeline inside the tracked task; deterministic mode is labeled honestly. There is no synthetic host skill engine.
• Reliable start/cancel: one active build is allowed per trusted scope. Durable task IDs survive reloads, stale owners recover safely, cancellation stops owned CLI children, and every terminal/error path restores the neuron page instead of leaving starting / 0% spinning.
• Unified governance: canonical rules, memory deduplication, compaction, knowledge references, and bounded context injection preserve identity, policy, priority, audience, scope, provenance, and evidence. Graphify stays integrated as an optional metadata provider behind MemoryGuard CodeGraph; it is not a separate MemoryGuard package.
• History, Governance, and Knowledge UI repaired: raw History sessions now open through the real SafeBridge/native V2 path, governance activity shows the responsible Agent instead of Unknown Agent, and /knowledge is again a full V2 bookshelf/detail experience instead of a JSON debug page.
• Clearer governance visualization: dominant automatic decisions collapse into readable groups, while the neuron graph uses a tighter level-aware layout and root-outward soft signal bands instead of projectile-like particles.

Final local regression: 1884 passed / 0 failed across 205 test files. See the v0.7.1 release record.

What's New in v0.7.0 (V2-only; published 2026-08-12)

v0.7.0 is V2-only. Local release acceptance passed and the release was published to GitHub and PyPI on 2026-08-12. MemoryGuard owns the CodeGraph boundary; Graphify is an optional extraction provider, not a second MemoryGuard runtime or a separate MemoryGuard package. The boundary and evidence below describe the release. The Graphify result is a real full-repository export/projection result, not a claim that upstream Graphify's full-repository test suite passed.

• V1 runtime physically retired: production entrypoint import closure has no V1 runtime/store modules. V1_ACTIVE is a migration starting state, not a runnable fallback. Legacy formats are readable only under memoryguard.migration; every other entrypoint fails closed with v2_upgrade_required. V1 data and migration backups remain rollback/audit evidence, never V2 runtime write targets.
• V2 control planes: Memory, Evidence, History, Source, Binding, and Group are V2-native boundaries. Memory atoms and evidence/decision receipts are separate from raw conversation History; authorized Source files/folders are separate from runtime state; trusted Agent Bindings select the governing Group.
• Canonical governance: canonical reconciliation folds rules into shared_baseline, agent_overlay, and project_overlay bundles, keeps durable source links, verifies parity, activates the canonical read path, then shadows old duplicates for recovery. V2 automatic organization performs exact/semantic duplicate detection inside one share group and records deduplicated, superseded, conflicted, or quarantined outcomes. Rule duplicate scans produce governed merge proposals; merge and supersede decisions keep evidence, scope, idempotency, and undo receipts.
• Cross-agent same-group governance: all members of one trusted share_group_id participate in the same bounded candidate and governance view; another group cannot enter it. Agent identity remains in provenance.
• Knowledge files and folders: a selected folder becomes a governed book and selected files become traceable documents. Content Plane owns source bodies; Knowledge stores metadata/references, supports re-ingest, remove/restore/ purge, and requires explicit review before memory candidates are accepted.
• GUI Agent and Group control: the native GUI discovers Agent names and instances, records source selections, lists bindings, binds members to shared or personal groups, checks drift, and can leave or dissolve a group. Group changes commit receipts and system outbox events transactionally.
• GUI builds and process cleanup: projection, Knowledge, import, history, maintenance, release, and compatibility work use durable V2 TaskRun status. Status survives reload, cancellation is cooperative and bounded, and owned background workers/process cleanup must finish before shutdown.
• CodeGraph / Graphify: MemoryGuard owns the trusted, body-free CodeGraph adapter and projection. Graphify is an optional extraction provider that supplies metadata-only exports; it is not a separate MemoryGuard runtime or package. CodeGraph preserves source role, provenance, source maps, revisions, tombstones, and outbox state, and exposes bounded query/path/explain/affected operations with production-only filtering.
• Security and rollback: unknown/corrupt state, missing scope, invalid provenance, reparse paths, unsafe metadata, and stale idempotency fail closed. Public receipts redact source bodies and paths; governance/audit/outbox records retain decisions. Release rollback restores a Content Plane blob and held occurrence through a scoped receipt rather than trusting an unbound backup path.
• Local release acceptance evidence: 1761 / 1761, with no skip or xfail; V1 retirement + CodeGraph 15 / 15; Graphify focused checks 3 / 3; canonical reconciliation ACCEPTED; RuleMerge 46 / 46; v3.2 27 / 27. The real full-repository Graphify export/projection covered 486 files / 11672 nodes / 38714 edges → 11667 canonical symbols / 38714 edges; query/path/ affected passed and failure atomicity was 0 throughout.
• Final packaging evidence: clean wheel 206 files, legacy bad=0; isolated package, CLI, and MCP all reported 0.7.0; desktop help passed.

Local release acceptance passed. v0.7.0 was published to GitHub and PyPI on 2026-08-12. These Graphify results cover the focused checks and the real full-repository export/projection only; they do not claim upstream Graphify's full-repository tests passed. See the v0.7.0 release record.

v0.6.2 compatibility baseline

The Python 3.10 SQLite correction remains part of the v0.7.0 upgrade baseline: Memory, Evidence, and Content schema preflights inspect a private copy of the SQLite main file plus -wal/-shm companions, and physical no-write checks do not observe or checkpoint the live database. The historical release note is preserved at docs/releases/v0.6.2.md.

Major V2 refactor in v0.6.0

v0.6.0 was a production data-plane refactor, not a storage-only upgrade:

• Authoritative V2 domains: Memory, Rules, Evidence, Content, Runtime, Projection, Assets, CodeGraph, Skills, and System state are separated into explicit SQLite domains with governed boundaries.
• Explicit cutover: V1_ACTIVE → V2_BUILDING → V2_READY → V2_ACTIVE is fail-closed; V2 never silently falls back to legacy stores or dual-writes after READY/ACTIVE.
• Lossless migration: frozen-source preparation uses coherent SQLite online backups, validates source/target evidence, rechecks live-source drift, and preserves V1 data plus migration backups for rollback.
• Native routing: MCP, CLI, GUI, and Hook surfaces are classified explicitly; the release closed the 233-surface cutover with 138 implemented routes, 95 retired routes, and zero neutral/blocker routes.
• Governed intelligence: Rule lifecycle and RuleMerge, extraction/enrichment, External MCP import, provider control-plane, conversation history, Knowledge Library, and GUI governance all use the V2 evidence and decision paths.
• Operational evidence: Reference Audit, per-domain SQLite health, guarded maintenance, rollback evidence, and safe unbound diagnostics are part of readiness and operations.

Why MemoryGuard

Persistent memory solves storage. It does not solve governance.

When several coding agents write into the same context, records become duplicated, stale, contradictory, over-broad, or unsafe to reuse. MemoryGuard sits between coding agents and their shared memory to keep that context usable.

System architecture

%%{init: {"theme":"base","themeVariables":{"background":"#071521","fontFamily":"Arial, sans-serif","fontSize":"14px","primaryTextColor":"#EEF4F8","lineColor":"#557287","edgeLabelBackground":"#071521","clusterBkg":"#0A1A29","clusterBorder":"#27445A"},"flowchart":{"htmlLabels":true,"curve":"basis","nodeSpacing":32,"rankSpacing":48,"padding":14}}}%%
flowchart TB
    Hosts["CODING-AGENT HOSTS<br/>Claude Code · Codex · Cursor · TRAE&nbsp;&nbsp;&nbsp;&nbsp;"]:::host
    Gateway["LOCAL INTEGRATION<br/>MCP stdio · redirect rules · lifecycle hooks&nbsp;&nbsp;&nbsp;&nbsp;"]:::gateway

    subgraph Core["GOVERNANCE CORE&nbsp;&nbsp;&nbsp;&nbsp;"]
        direction LR
        Identity["TRUST<br/>identity · scope&nbsp;&nbsp;&nbsp;&nbsp;"]:::core
        MemoryAPI["MEMORY<br/>governed I/O&nbsp;&nbsp;&nbsp;&nbsp;"]:::active
        Rules["RULES<br/>scope · assignment&nbsp;&nbsp;&nbsp;&nbsp;"]:::rule
        HistoryAPI["HISTORY<br/>search · timeline&nbsp;&nbsp;&nbsp;&nbsp;"]:::history
        Security["SAFETY<br/>validate · quarantine&nbsp;&nbsp;&nbsp;&nbsp;"]:::danger

        Identity --> MemoryAPI
        Identity --> Rules
        Identity --> HistoryAPI
        MemoryAPI --> Security
    end

    subgraph Stores["LOCAL GOVERNED STORES&nbsp;&nbsp;&nbsp;&nbsp;"]
        direction LR
        SharedDB[("V2 DOMAIN STORES<br/>Memory · Rules · Evidence · Content&nbsp;&nbsp;&nbsp;&nbsp;")]:::store
        HistoryDB[("HISTORY STORE<br/>isolated conversations&nbsp;&nbsp;&nbsp;&nbsp;")]:::historyStore
        AuditDB[("RECOVERY STORE<br/>versions · receipts · backups&nbsp;&nbsp;&nbsp;&nbsp;")]:::store
    end

    Bootstrap["BOUNDED CONTEXT BOOTSTRAP<br/>mandatory rule pack · relevant recall&nbsp;&nbsp;&nbsp;&nbsp;"]:::bootstrap
    Control["HUMAN CONTROL<br/>CLI · desktop governance console&nbsp;&nbsp;&nbsp;&nbsp;"]:::surface

    Hosts --> Gateway --> Identity
    MemoryAPI --> SharedDB
    Rules --> SharedDB
    HistoryAPI --> HistoryDB
    Security --> AuditDB
    SharedDB --> Bootstrap
    Control --> Identity

    classDef host fill:#12243A,stroke:#38D5C8,color:#EEF4F8,stroke-width:1.4px;
    classDef gateway fill:#0D3338,stroke:#38D5C8,color:#EEF4F8,stroke-width:2.4px;
    classDef core fill:#12243A,stroke:#557287,color:#EEF4F8,stroke-width:1.4px;
    classDef active fill:#0D383A,stroke:#38D5C8,color:#EEF4F8,stroke-width:2px;
    classDef rule fill:#3B2C18,stroke:#F3B562,color:#EEF4F8,stroke-width:1.8px;
    classDef history fill:#102F45,stroke:#73C7F5,color:#EEF4F8,stroke-width:1.8px;
    classDef danger fill:#3A2028,stroke:#EA6A6A,color:#EEF4F8,stroke-width:1.8px;
    classDef bootstrap fill:#EEF4F8,stroke:#38D5C8,color:#071521,stroke-width:2.4px;
    classDef store fill:#0B1624,stroke:#7F96A8,color:#EEF4F8,stroke-width:1.4px;
    classDef historyStore fill:#102436,stroke:#73C7F5,color:#EEF4F8,stroke-width:1.4px;
    classDef surface fill:#EEF4F8,stroke:#38D5C8,color:#071521,stroke-width:2px;

    style Core fill:#081827,stroke:#27445A,stroke-width:1px,color:#EEF4F8
    style Stores fill:#081827,stroke:#27445A,stroke-width:1px,color:#EEF4F8
    linkStyle default stroke:#557287,stroke-width:1.4px;

Quick start

1. Install

python -m pip install agent-memguard

For the desktop governance console:

python -m pip install "agent-memguard[gui]"

2. Authorize the current project

memoryguard source add .

3. Connect or repair your coding agent

Global provider configuration is rebuilt from the real binding in the canonical user data home. The command is idempotent and removes superseded MemoryGuard project-level overrides after a successful global takeover.

# Repair one provider
memoryguard provider repair claude
memoryguard provider repair codex
memoryguard provider repair cursor
memoryguard provider repair trae

# Repair every detected provider
memoryguard provider repair all

Restart the host after installation, then verify the integration:

memoryguard doctor
memoryguard mcp-status
memoryguard hooks status --provider all

Launch the desktop console:

memoryguard gui

memoryguard-gui . remains available for desktop shortcuts. A bare memoryguard gui always opens the canonical user-level control directory (default %LOCALAPPDATA%\MemoryGuard on Windows), so running it from a project or from C:\Windows\System32 cannot silently switch databases. MEMORYGUARD_WORKSPACE is an explicit operator override; an explicit memoryguard gui <project-path> or memoryguard gui --workspace <project-path> selects a specific workspace. It does not remember a previously selected project or open a folder picker. On Windows, memoryguard gui detaches the native window from the terminal, so closing PowerShell does not close the GUI.

Provider-specific setup and behavior:

• Claude Code installation
• Codex installation
• Cursor installation

Stable Codex / Router binding

Codex/Router binds MemoryGuard to the stable local Codex program and control installation. An account profile is an endpoint/alias, not a new memory owner: switching profiles automatically discovers or repairs the profile and reuses the verified Agent binding and active group. Request identity remains fail-closed; this does not share records across machines or with arbitrary accounts.

Upgrade

MemoryGuard currently upgrades through Python's package manager:

python -m pip install --upgrade agent-memguard
memoryguard --version
memoryguard doctor

If you installed the GUI extra, keep it during the upgrade:

python -m pip install --upgrade "agent-memguard[gui]"

There is no package self-update command. The package manager is the authoritative package-upgrade path; memoryguard upgrade below is the explicit workspace migration flow, not a package updater.

Upgrade an existing V1 data home

Upgrade the package, then run the verified migration. No workspace, data-home, apply, or confirmation arguments are required for the normal user-level data home:

python -m pip install --upgrade agent-memguard
memoryguard --version                    # 0.7.7
memoryguard upgrade
memoryguard doctor

The command prepares V2, validates the frozen and live source evidence, migrates Agent/Group control, activates only after all gates pass, and removes only the backup batch belonging to that successful migration. Re-running it on V2_ACTIVE is idempotent. For a zero-write report, use:

memoryguard upgrade --preview

Advanced explicit workspace/data-home options remain available for operators managing an isolated installation. A failed gate stays non-active and preserves its evidence; successful activation does not keep a redundant migration backup.

Existing pre-V2 workspaces: explicit V2 cutover

v0.6.0 never auto-activates an existing workspace. Upgrade the package first, then use the packaged operator CLI:

# Read-only manifest status
memoryguard-v2 status -w .

# Build a frozen-source V2 shadow and stop at V2_READY
memoryguard-v2 prepare -w . --apply

# Activate only after the prepare result is V2_READY / ready=true
memoryguard-v2 activate -w . --confirm V2_ACTIVE

The prepare step uses coherent SQLite online backups, preserves V1 and migration-backups, and rechecks live-source drift before READY. Activation performs another fresh drift check before changing the manifest. Do not delete legacy V1 data or migration backups as part of the upgrade.

Knowledge Library

The desktop console can turn a selected folder or file set into one governed local knowledge library. Source files remain where they are; MemoryGuard stores the searchable index in its user data home instead of copying a runtime database into every source project. Knowledge metadata never becomes a second source-body store.

Open the desktop console and choose Knowledge Library. Remote embedding or model-backed indexing is opt-in and requires explicit authorization; local full-text retrieval remains available without sending source text to a remote provider.

CodeGraph refresh

The first CodeGraph build is an explicit, confirmed full build. After a scope has been built, each successful trusted file write can trigger an incremental refresh for that scope, subject to strict source-path and active-binding validation. Unchanged content hashes are a no-op; deleted files are retired; the next context receives one bounded affected receipt. MemoryGuard does not run a daemon or watcher for this path and does not infer paths from shell or free-form text.

Desktop console surfaces

The GUI follows a seven-page information architecture:

1. Governance Overview
2. Data Sources & Agents
3. Memory Core
4. CodeGraph
5. Rules & Habits
6. Conversation History
7. Risk Signals & Governance Console

Agent lists use readable program/provider names; the underlying ID remains available in the detail view. Empty data is shown as an explicit empty state.

Write and governance lifecycle

%%{init: {"theme":"base","themeVariables":{"background":"#071521","fontFamily":"Arial, sans-serif","fontSize":"14px","primaryTextColor":"#EEF4F8","lineColor":"#557287","edgeLabelBackground":"#071521","clusterBkg":"#0A1A29","clusterBorder":"#27445A"},"flowchart":{"htmlLabels":true,"curve":"basis","nodeSpacing":30,"rankSpacing":42,"padding":14}}}%%
flowchart TD
    subgraph Intake["01 · INTAKE    "]
        direction LR
        Write(["Memory write    "]):::entry
        Scope["Resolve identity<br/>scope · audience&nbsp;&nbsp;&nbsp;&nbsp;"]:::core
        Validate{"Authorized?&nbsp;&nbsp;&nbsp;&nbsp;"}:::decision
        Reject["Reject<br/>no persistence&nbsp;&nbsp;&nbsp;&nbsp;"]:::danger
        Write --> Scope --> Validate
        Validate -- NO --> Reject
    end

    subgraph Organize["02 · ORGANIZE&nbsp;&nbsp;&nbsp;&nbsp;"]
        direction TB
        Secret{"Sensitive?&nbsp;&nbsp;&nbsp;&nbsp;"}:::decision
        Quarantine["Quarantine<br/>outside active set&nbsp;&nbsp;&nbsp;&nbsp;"]:::danger
        Compare["Classify · compare<br/>governed records&nbsp;&nbsp;&nbsp;&nbsp;"]:::active
        Relation{"Relationship&nbsp;&nbsp;&nbsp;&nbsp;"}:::decision
        New["NEW<br/>create active record&nbsp;&nbsp;&nbsp;&nbsp;"]:::result
        Duplicate["DUPLICATE<br/>merge provenance&nbsp;&nbsp;&nbsp;&nbsp;"]:::result
        Correction["CORRECTION<br/>supersede old record&nbsp;&nbsp;&nbsp;&nbsp;"]:::rule
        Conflict["CONFLICT<br/>preserve both sides&nbsp;&nbsp;&nbsp;&nbsp;"]:::danger

        Secret -- YES --> Quarantine
        Secret -- NO --> Compare --> Relation
        Relation --> New
        Relation --> Duplicate
        Relation --> Correction
        Relation --> Conflict
    end

    subgraph Govern["03 · GOVERN&nbsp;&nbsp;&nbsp;&nbsp;"]
        direction LR
        Receipt[("Evidence event<br/>version receipt&nbsp;&nbsp;&nbsp;&nbsp;")]:::store
        Review["CLI or desktop review&nbsp;&nbsp;&nbsp;&nbsp;"]:::surface
        Action["Correct · merge<br/>restore · delete&nbsp;&nbsp;&nbsp;&nbsp;"]:::rule
        Snapshot["Reversible<br/>snapshot&nbsp;&nbsp;&nbsp;&nbsp;"]:::active
        Receipt --> Review --> Action --> Snapshot
    end

    Validate -- YES --> Secret
    Quarantine --> Receipt
    New --> Receipt
    Duplicate --> Receipt
    Correction --> Receipt
    Conflict --> Receipt

    classDef entry fill:#EEF4F8,stroke:#38D5C8,color:#071521,stroke-width:2.4px;
    classDef core fill:#12243A,stroke:#557287,color:#EEF4F8,stroke-width:1.5px;
    classDef decision fill:#0D3338,stroke:#38D5C8,color:#EEF4F8,stroke-width:2px;
    classDef active fill:#0D383A,stroke:#38D5C8,color:#EEF4F8,stroke-width:2px;
    classDef result fill:#12243A,stroke:#38D5C8,color:#EEF4F8,stroke-width:1.6px;
    classDef rule fill:#3B2C18,stroke:#F3B562,color:#EEF4F8,stroke-width:1.8px;
    classDef danger fill:#3A2028,stroke:#EA6A6A,color:#EEF4F8,stroke-width:1.8px;
    classDef store fill:#0B1624,stroke:#7F96A8,color:#EEF4F8,stroke-width:1.4px;
    classDef surface fill:#EEF4F8,stroke:#38D5C8,color:#071521,stroke-width:2px;

    style Intake fill:#081827,stroke:#27445A,stroke-width:1px,color:#EEF4F8
    style Organize fill:#081827,stroke:#27445A,stroke-width:1px,color:#EEF4F8
    style Govern fill:#081827,stroke:#27445A,stroke-width:1px,color:#EEF4F8
    linkStyle default stroke:#557287,stroke-width:1.4px;

The console is not an approval queue. Agents keep moving. MemoryGuard records the outcome and exposes the evidence needed to correct it later.

What you can govern

Rules and history stay separate

MemoryGuard deliberately keeps governed long-term memory and raw conversation history on different paths.

History retrieval is progressive: search results, then a bounded timeline, then an explicitly selected turn or session. Extracting from history creates a preview first; it does not silently write a long-term memory.

%%{init: {"theme":"base","themeVariables":{"background":"#071521","fontFamily":"Arial, sans-serif","fontSize":"14px","primaryTextColor":"#EEF4F8","lineColor":"#557287","edgeLabelBackground":"#071521","clusterBkg":"#0A1A29","clusterBorder":"#27445A"},"flowchart":{"htmlLabels":true,"curve":"basis","nodeSpacing":30,"rankSpacing":42,"padding":14}}}%%
flowchart LR
    subgraph HistoryPath["CONVERSATION EVIDENCE    "]
        direction TB
        Archive[("Raw local history    ")]:::historyStore
        Search["Search summaries    "]:::history
        Timeline["Bounded timeline    "]:::history
        Read["Explicit turn or session    "]:::history
        Preview["Evidence-backed<br/>extraction preview&nbsp;&nbsp;&nbsp;&nbsp;"]:::history
        Confirm["Explicit acceptance&nbsp;&nbsp;&nbsp;&nbsp;"]:::surface
        Isolation["NO AUTOMATIC<br/>BOOTSTRAP PATH&nbsp;&nbsp;&nbsp;&nbsp;"]:::barrier

        Archive --> Search --> Timeline --> Read --> Preview --> Confirm
        Archive -.-> Isolation
    end

    subgraph GovernedMemory["GOVERNED LONG-TERM MEMORY&nbsp;&nbsp;&nbsp;&nbsp;"]
        direction TB
        Mandatory["Scoped mandatory rules&nbsp;&nbsp;&nbsp;&nbsp;"]:::rule
        Assignments["Agent · project<br/>role · group scope&nbsp;&nbsp;&nbsp;&nbsp;"]:::core
        RulePack["Mandatory-rule<br/>budget&nbsp;&nbsp;&nbsp;&nbsp;"]:::budget
        Ordinary["Facts · preferences<br/>projects · procedures&nbsp;&nbsp;&nbsp;&nbsp;"]:::memory
        Recall["Task-relevant<br/>recall budget&nbsp;&nbsp;&nbsp;&nbsp;"]:::budget
        Context["BOUNDED CONTEXT PACKET&nbsp;&nbsp;&nbsp;&nbsp;"]:::context

        Mandatory --> Assignments --> RulePack --> Context
        Ordinary --> Recall --> Context
    end

    HistoryPath ==>|GOVERNED WRITE&nbsp;&nbsp;&nbsp;&nbsp;| GovernedMemory

    classDef rule fill:#3B2C18,stroke:#F3B562,color:#EEF4F8,stroke-width:1.8px;
    classDef core fill:#12243A,stroke:#557287,color:#EEF4F8,stroke-width:1.4px;
    classDef memory fill:#0D383A,stroke:#38D5C8,color:#EEF4F8,stroke-width:1.8px;
    classDef budget fill:#12243A,stroke:#38D5C8,color:#EEF4F8,stroke-width:1.6px;
    classDef context fill:#EEF4F8,stroke:#38D5C8,color:#071521,stroke-width:2.4px;
    classDef history fill:#102F45,stroke:#73C7F5,color:#EEF4F8,stroke-width:1.6px;
    classDef historyStore fill:#102436,stroke:#73C7F5,color:#EEF4F8,stroke-width:1.6px;
    classDef surface fill:#EEF4F8,stroke:#73C7F5,color:#071521,stroke-width:2px;
    classDef barrier fill:#3A2028,stroke:#EA6A6A,color:#EEF4F8,stroke-width:2px;

    style GovernedMemory fill:#081827,stroke:#27445A,stroke-width:1px,color:#EEF4F8
    style HistoryPath fill:#081827,stroke:#27445A,stroke-width:1px,color:#EEF4F8
    linkStyle default stroke:#557287,stroke-width:1.4px;

Supported hosts

Provider status is reported honestly as redirected, observed, operational, or unsupported. MemoryGuard does not claim it can disable every host's native memory when the host exposes no reliable integration point.

Architecture

V2 uses separate authoritative SQLite domains rather than one shared-memory database. The runtime reads and writes V2 only after the manifest reaches V2_ACTIVE; V2_BUILDING and V2_READY never silently fall back or dual-write. Evidence remains traceable without being treated as automatically trusted memory.

Privacy and safety

• MemoryGuard runs as a local MCP stdio server.
• All governed data stays local unless you explicitly authorize a remote model or embedding operation.
• The Knowledge Library database uses MEMORYGUARD_HOME or the platform user data directory, so a selected source folder does not receive its own knowledge database.
• V2 authoritative workspace state is separated under .memoryguard/ into explicit Memory, Rules, Evidence, Content, Runtime, Projection, Assets, CodeGraph, Skills, and System domains; History, Source, Binding, and Group control are V2-native surfaces. Legacy V1 artifacts are preserved as local rollback/audit evidence after cutover and are no longer the active V2 runtime write path; only memoryguard.migration may read them.
• Source scanning is read-only by default.
• Mutating governance paths use validation, explicit scope, provenance, and reversible state.
• Quarantined records stay outside active shared memory.
• Raw conversation history is never injected into bootstrap automatically.
• Shared-group history access follows current active membership and does not grant deletion rights over another Agent's source.

CLI

The installed memoryguard command exposes these top-level operations:

The old V1 plan, apply, verify, undo, import, and gc workflows may remain parseable as explicit retired compatibility surfaces, but are not a V1 runtime path. Under V2_ACTIVE they return a stable retired result instead of writing through a legacy store. Legacy data input is accepted only by the explicit memoryguard.migration upgrade flow.

Run memoryguard --help or memoryguard <command> --help for the live command reference.

MCP API

The MCP server exposes tools for:

• governed memory read, search, write, update, delete, and status;
• bounded context bootstrap with mandatory-rule isolation;
• rule creation, feedback, merge governance, undo, and scope statistics;
• Agent binding and shared-group inspection;
• source scanning, graph projection, import previews, and build planning;
• external MCP discovery and import;
• document extraction previews and candidate acceptance;
• conversation-history search, timeline, explicit read, export, deletion, and extraction preview;
• provider installation and host-agent enrichment.

Use MCP tools/list as the source of truth for the exact tool set supported by the installed version.

Project links

• PyPI package
• GitHub releases
• Changelog
• v0.7.6 release record
• v0.7.5 release record
• v0.7.4 release record
• v0.7.3 release record
• v0.7.2 release record
• v0.7.1 release record
• v0.7.0 release gate
• Memory continuity and lossless storage spec
• Contributing guide
• Contributor License Agreement
• Issue tracker

Roadmap

• Current release line: v0.7.7 makes bare provider repair safe in a verified, uniquely bound control home and aligns installed Codex MCP/Hook repairs to the current interpreter while preserving Agent and shared-group identity. v0.7.6 makes Codex Hook/MCP runtime selection consistent through one immutable snapshot, shortens Hook state lock windows for concurrent tools, and keeps bootstrap success/failure state honest with explicit mandatory-overflow fail-closed handling. v0.7.5 makes conflict review self-contained and fail-closed with redacted member snapshots, localized reasons, invalid-candidate protection, and separate historical/actionable counts. v0.7.4 unified canonical governance updates, bounded mandatory context warnings, readable governance labels, and immutable Codex MCP snapshots. The v0.7.3 shared-group history recall, v0.7.2 write/read scope, Codex Hook lifecycle and transport, and Python 3.10 compatibility corrections. The v0.7.1 V2-only migration and desktop lifecycle work remains documented as historical release context.
• Acceptance boundary: the Graphify evidence is the focused 3 / 3 result plus the real full-repository export/projection described above. It does not claim that upstream Graphify's full-repository test suite passed.
• Next after release: broader CodeGraph/Skills ingestion, more operator-friendly maintenance reports, and additional migration observability. Long-term records are not retired merely because they are old.
• Later: team and enterprise capabilities only after validated demand.

Contributing

Issues and pull requests are welcome. Read CONTRIBUTING.md before submitting a change. Pull requests require agreement to the CLA.

License

MIT