@jungjaehoon/mama-os 0.53.0 → 0.57.0

package/CHANGELOG.md

@@ -7,6 +7,70 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 ## [Unreleased]
 
+## [0.57.0] - 2026-09-27
+
+The product layer is rebuilt around one owner agent on Claude or Codex; see the root CHANGELOG
+and docs/start/owner-setup.md. Breaking: the CLI is `init | secret | daemon | replay | status |
+stop`, the Telegram bot token moves from config.yaml to auth.env (MAMA_TELEGRAM_TOKEN), and keys of
+the old product layer are ignored.
+
+## [0.56.0] - 2026-09-13
+
+### Added
+
+- Owner messages, completed responses, and connector deltas carry immutable observation references
+  through durable sessions and the same model run.
+- `registry_correct` is available to the owner agent through native and Code-Act catalogs with
+  signed scope, principal, agent, and revision enforcement.
+- Raw observation detail and graph current-projection responses preserve exact evidence identity.
+
+## [0.55.0] - 2026-09-13
+
+### Added
+
+- Registry lookup/upsert are available through the gateway and Code-Act catalogs.
+- Gateway and retained operational HTTP saves carry explicit record item and actor references.
+- Owner-action receipts support truthful service operation origins without synthetic model runs.
+
+## [0.54.0] - 2026-09-12
+
+### Removed
+
+- Removed the second localhost runtime on port 3849, including startup/takeover, shutdown cleanup,
+  health probing, WebSocket proxying, direct viewer-session routes, and query-token authentication.
+- Unauthorized requests no longer parse or record query-token-specific metadata.
+- Removed direct `ws` runtime and type dependencies.
+
+### Changed
+
+- Runtime status awaits live health checks and gives live gateway pass/fail state precedence over
+  stored connector settings while limiting generic output to the public connector catalog.
+
+### Kept
+
+- The port 3847 operational API, reports, raw evidence, graph, uploads, health, messenger delivery,
+  durable session storage, and shutdown ordering.
+
+## [0.53.4] - 2026-09-12
+
+### Removed
+
+- Removed the unused browser Viewer, frontend workspace, static routes, and UI command queue.
+- Removed the active `webchat_send` tool advertisement and executor while retaining historical
+  stored-effect compatibility.
+
+### Kept
+
+- The shared operational API on port 3847, graph routes, report delivery, messenger and native
+  tool paths, CLI commands, and durable session continuity remain supported.
+
+## [0.53.3] - 2026-09-12
+
+### Changed
+
+- Connector collection and raw evidence readers now use the shared core source archive while
+  preserving existing revision, provenance, isolation, and reopen behavior.
+
 ## [0.51.2] - 2026-09-08
 
 ### Fixed

package/README.md

@@ -1,267 +1,70 @@
-# @jungjaehoon/mama-os
+# MAMA OS
 
-> Bounded, provenance-backed working context for AI agents running on your machine.
+One owner agent on Claude or Codex watches connected work, keeps task revisions and evidence,
+answers on Telegram, publishes reports and a board, and recalls corrections.
+[mama-core](../mama-core/README.md) supplies the shared engine.
 
-## The Problem
+Current manifest: **0.57.0**. This README describes the unreleased `rebuild/owner-flow` source.
+Use [owner setup](../../docs/start/owner-setup.md) with Node.js 22.13+ and pnpm.
 
-Your knowledge is everywhere — Slack threads, email chains, code reviews, meeting notes, spreadsheets, Telegram messages. No human can track all of it. Important decisions get buried. Context gets lost between tools. When you need to make a decision, the information that would help is scattered across ten different apps and three months of history.
+## Start and operate
 
-This isn't just a memory problem. It's a bounded context problem. You don't just need to
-_store_ information — you need something that reads everything, connects the dots, identifies what
-matters, proves where it came from, and keeps agents inside the scope they were given.
-
-## What MAMA OS Does
-
-MAMA OS is a local AI runtime that connects to your apps, reads continuously, and turns scattered
-records into scoped, auditable context for agents and humans.
-
-**What the agents actually do:**
-
-- **Identify what matters** — Out of thousands of daily messages, surface the decisions, deadlines, and changes that affect your work
-- **Connect across sources** — A Slack conversation + a Trello card + an email attachment about the same project are linked automatically
-- **Track decision evolution** — Not just what was decided, but what it replaced, what it builds on, and what it contradicts
-- **Operate inside envelopes** — Gateway and worker calls carry signed scope boundaries and audit rows
-- **Preserve provenance** — Memory writes can point back to source refs, model runs, tool traces, and envelope hashes
-- **Search with evidence** — Strict memory search can reject vector-only noise and show which lexical, entity, scope, or graph signals confirmed a result
-- **Compile actionable knowledge** — Promoted decisions become an Obsidian wiki: an append-only daily journal plus durable lesson pages that strengthen with evidence
-- **Evolve their own triggers** — The operator loop authors triggers from recurring situations, fires them to recall the right memory, and scores them by whether delivered reports actually cite them
-- **Brief you proactively** — When you start working, relevant context from all sources is already there — you didn't ask for it
-
-```
-Without MAMA:  You read 5 Slack channels, 3 email threads, check Trello,
-               re-read old PRs, then try to piece together the full picture.
-
-With MAMA:     Agents already read everything. You get a briefing with
-               what changed, what's at risk, and what needs your decision.
-```
-
-**This is what local AI agents should do** — read every channel, every thread, every document, every
-day, then explain exactly which evidence they used and which permission boundary they were inside.
-
-- **Private by design** — All data stays on your device. Nothing leaves your machine.
-- **AI-independent** — Works with Claude, Codex, or any future backend. Your memory outlives any AI provider.
-
-## How It Runs
-
-MAMA OS runs AI agents through each backend's **supported local runtime path**: Claude CLI,
-Codex app-server, or Cline's official Hub runtime.
-
-Some third-party agent frameworks (OpenClaw, etc.) use unofficial API access, token extraction, or header spoofing — approaches that violate provider policies and risk account suspension. MAMA OS doesn't do any of that. If you have Claude Code, Codex CLI, or Cline CLI installed and authenticated, MAMA OS uses that backend's supported local runtime path. No token extraction or header spoofing is required.
+From the repository root:
 
 ```bash
-# Already authenticated Claude, Codex, or Cline?
-mama --help
-mama status   # follow the reported next action until complete
+pnpm install
+pnpm build
+node packages/standalone/dist/cli/index.js init
 ```
 
-## How It's Secured
-
-MAMA OS has full system access — so security is not optional, it's foundational.
-
-- **Local-only by default** — Binds to localhost. External access requires explicit tunnel setup with authentication (Cloudflare Zero Trust).
-- **Signed runtime envelopes** — Gateway and worker tool calls carry verifiable scope, expiry, and
-  actor context before irreversible side effects are allowed.
-- **Provenance ledger** — Memory writes, raw refs, model runs, and tool traces can be audited after
-  the fact without exposing prompt bodies or hidden connector payloads.
-- **5-layer prompt injection defense** — Output sanitization, channel trust boundaries, silent mode for unknown sources, bulk extraction limits. Built from a real incident, not theory.
-- **Intrusion detection** — Honeypot traps for scanner probes (`.git`, `.env`, `wp-login.php`), per-IP suspicion scoring, automatic tarpit delays, and IP deny-listing when thresholds are exceeded.
-- **Agent permission tiers** — Tier 1 (full access), Tier 2 (read + memory write), Tier 3 (read-only). Each agent only gets the tools it needs.
-- **Owner console (v0.22+)** — the `owner_console` role is granted ONLY in an allowlisted telegram chat's 1:1 DM (`telegram.allowed_chats` is the trust anchor). It reads operational artifacts (`board_read`, `audit_findings_read`), can browse/download/upload Drive files, OCR and overlay translated image text, send files back to the current Telegram chat, and issue work (`report_request`) fire-and-forget. The verified owner gets the complete Drive composition surface and may upload to the folder selected in the active request even when it is not a configured deliverable root. Configured-root capabilities remain available and are validated when supplied; non-owner Drive operations require role permission and configured connector/envelope scope and cannot select arbitrary roots. Uploads and outbound files accept only regular non-symlink files in the private MAMA workspace; memory writes refuse secret-shaped content. On a clean installation, prepare the isolated OCR runtime with `pnpm setup:ocr` and verify it with `pnpm check:ocr`; `MAMA_OCR_PYTHON` can select an equivalent managed runtime. Translated overlays also require a Korean/CJK font: macOS uses Apple SD Gothic Neo, while Ubuntu/Debian should install `fonts-noto-cjk`; other systems can set `MAMA_KOREAN_FONT` to a readable `.ttf`/`.ttc` file. The setup check verifies this dependency instead of failing on the first real overlay. DOCX/XLSX extraction runs in a memory- and time-bounded child process. PDF extraction uses PDFKit on macOS and requires Poppler's `pdftotext` on Linux (`apt install poppler-utils` or the distribution equivalent).
-- **Stage-2 workorder pipeline (v0.28+, always on)** — scheduled board/wiki/memory-promotion runs become durable, occurrence-keyed workorders consumed serially on the operator lane. Since v0.41.0 (One MAMA) every scheduled turn runs as the same `owner_console` principal as chat, with the one console brief plus a host-authored turn-kind section; the host projects the grant (artifact tools added; administration, sends, uploads and per-kind mutation blocked) and renders the board pipeline slot itself. Every worker treats connector evidence as untrusted data rather than instructions. The native ledger remains authoritative for owner tasks and pipeline state. `MAMA_STAGE2_WORKORDERS` is retired: leave it unset (or set `on` for compatibility); explicit `off` or `shadow` values fail startup.
-- **Verified temporal effects (opt-in)** — `MAMA_TEMPORAL_RECONCILE=on` requires the always-on workorder consumer, a Claude, Codex, or Cline backend, the trusted `task_temporal_reconcile` tool, and a working worker transport. It scans all open scheduled owner rows once per minute before applying bounded admission caps, and exposes a separate temporal projection that never turns overdue into a workflow status (`closed` reflects terminal lifecycle). Fresh evidence may resolve, finalize, or defer one native owner-task occurrence. Task, generation, authoritative receipt, and workorder completion commit atomically; legacy receipts remain quarantined, every temporal gateway call rechecks active authority, and compiled raw evidence must stay bound to the task source. Shutdown durably pauses temporal attempts before waiting for worker drainage. Model prose is represented in audit rows only by length/SHA-256 references. Default `off` pauses temporal work and preserves existing behavior.
-- **Fail-safe shutdown** — When an intrusion cannot be contained, MAMA shuts itself down gracefully rather than operating in a compromised state.
-
-These aren't theoretical protections. The prompt injection defense was built after a real attack where an adversary injected a fake "server failure" message into a monitored channel, causing the AI agent to voluntarily expose system configuration. The IP banning system has blocked actual intrusion attempts in production.
-
-See the full [Security Guide](../../docs/guides/security.md) for Cloudflare Zero Trust setup, token authentication, threat scenarios, agent isolation, and Code-Act sandbox security.
-
-## Quick Start
-
-```bash
-# 1. Authenticate one backend (one-time)
-claude auth login   # or: codex login
-cline auth cline    # for the hosted Cline backend
-
-# 2. Install and follow the self-teaching contract
-npm install -g @jungjaehoon/mama-os
-mama --help
-mama status
-```
-
-Use `mama status --json` when another agent is performing setup. It reports ordered missing actions
-for initialization, Telegram owner anchoring, an authenticated work-source connector, daemon
-startup, and the first confirmed report. `mama gateway telegram --token-stdin` accepts the bot
-token without echoing it; `mama connector add <name>` handles non-Telegram work sources.
-
-**Prerequisites:** Node.js >= 22.13.0, one authenticated backend CLI (Claude, Codex, or Cline), 500MB disk space.
-
-## Connectors (13)
-
-MAMA connects to your apps and extracts structured facts into the memory graph.
-
-```bash
-mama connector add slack      # Activate + auth guide
-mama connector list           # Status of all connectors
-```
-
-| Connector       | Prerequisites                                                                                                                        | Config                           |
-| --------------- | ------------------------------------------------------------------------------------------------------------------------------------ | -------------------------------- |
-| **Slack**       | Bot Token (api.slack.com → OAuth scopes)                                                                                             | `bot_token`, `app_token`         |
-| **Discord**     | Bot Token (discord.com/developers → MESSAGE CONTENT INTENT)                                                                          | `token`, `default_channel_id`    |
-| **Telegram**    | Bot Token (@BotFather); text, captions, photos, image documents, and regular documents are accepted                                  | `token`, `allowed_chats`         |
-| **Chatwork**    | API Token (account settings)                                                                                                         | `api_token`, `room_ids`          |
-| **iMessage**    | macOS only (reads local chat.db)                                                                                                     | No config needed                 |
-| **Gmail**       | [gws CLI](https://github.com/nicholasgasior/gws) installed + Google OAuth                                                            | `gws` in PATH                    |
-| **Calendar**    | gws CLI installed + Google OAuth                                                                                                     | `gws` in PATH                    |
-| **Drive**       | gws CLI installed + Google OAuth; bounded polls continue through change-page backlogs, owner console also has direct file operations | `gws` in PATH                    |
-| **Sheets**      | gws CLI installed + Google OAuth                                                                                                     | `gws` in PATH, `spreadsheet_ids` |
-| **Notion**      | Integration Token (notion.so/my-integrations)                                                                                        | `api_token`, `database_ids`      |
-| **Obsidian**    | [Obsidian](https://obsidian.md) installed + [Obsidian Terminal](https://github.com/polyipseity/obsidian-terminal) plugin enabled     | `vault_path` in config.yaml      |
-| **Trello**      | API Key + Token (trello.com/app-key)                                                                                                 | `api_key`, `token`, `board_ids`  |
-| **Claude Code** | Claude Code plugin installed                                                                                                         | Automatic via hooks              |
-
-**Google Workspace connectors** (Gmail, Calendar, Drive, Sheets) require the [gws CLI](https://github.com/nicholasgasior/gws) — a Google Workspace command-line tool. Install it, run `gws auth` once for OAuth, then MAMA polls via CLI.
-
-Each connector classifies its source (truth / hub / spoke / reference) for the 3-pass extraction pipeline. Config: `~/.mama/connectors.json`.
-
-For connector polling and preconfigured deliverable roots, enable Drive and declare non-ignored
-channels. `folderId` scopes that channel to one folder; `driveId` scopes it to a shared drive. These
-roots can issue short-lived destination capabilities, but a verified owner-console request is not
-limited to them when it explicitly selects another Drive folder.
-
-```json
-{
-  "drive": {
-    "enabled": true,
-    "pollIntervalMinutes": 5,
-    "channels": {
-      "project-docs": {
-        "role": "deliverable",
-        "name": "Project docs",
-        "folderId": "GOOGLE_DRIVE_FOLDER_ID"
-      }
-    },
-    "auth": { "type": "cli", "cli": "gws", "cliAuthCommand": "gws auth login" }
-  }
-}
-```
-
-## Knowledge Agents
-
-MAMA OS runs specialized agents for knowledge management — not coding (that's what Claude Code does natively).
-
-| Agent               | Role                                                | Requires                                                                                              |
-| ------------------- | --------------------------------------------------- | ----------------------------------------------------------------------------------------------------- |
-| **Conductor**       | Orchestrates other agents, handles user chat        | —                                                                                                     |
-| **Dashboard Agent** | Generates project briefings from connected sources  | —                                                                                                     |
-| **Wiki Agent**      | Compiles knowledge into Obsidian vault              | [Obsidian](https://obsidian.md) + [Terminal plugin](https://github.com/polyipseity/obsidian-terminal) |
-| **Memory Agent**    | Extracts decisions from conversations automatically | —                                                                                                     |
-
-Agents delegate via `delegate()` with skill injection and automatic retry. Configure in `~/.mama/config.yaml`.
-
-## Viewer
-
-One web console at `http://localhost:3847/viewer` (`/` redirects there), opening on the
-operator board. PWA-enabled for mobile (add to home screen).
-
-Three groups:
-
-| Group         | View       | What it shows                                                                                    |
-| ------------- | ---------- | ------------------------------------------------------------------------------------------------ |
-| **Operator**  | Board      | Four agent-published report slots (briefing, action required, decisions, pipeline) live over SSE |
-| **Operator**  | Tasks      | The task board fed from your channels                                                            |
-| **Operator**  | Triggers   | The trigger loop's library with an owner veto tray                                               |
-| **Knowledge** | Memory     | Interactive reasoning graph (1000+ nodes), search, export                                        |
-| **Knowledge** | Wiki       | Knowledge base (syncs with Obsidian vault)                                                       |
-| **System**    | Runtime    | Read-only runtime truth: backend, model, gateways, health                                        |
-| **System**    | Connectors | Connector status and last poll                                                                   |
-| **System**    | Logs       | Daemon logs with filtering, pinning, stats, WebSocket and polling                                |
-
-Each view has its own hash route, e.g. `/viewer#operator/tasks?task=42`.
-
-## Gateway Integrations
-
-Run MAMA as a bot in Discord, Slack, Telegram, or Chatwork. Configure one gateway in
-`~/.mama/config.yaml`, then use `mama status` to verify its owner trust anchor.
-
-## Architecture
-
-```
-Connectors (13)              Gateways (4)
-Slack, Gmail, Sheets...      Discord, Slack, Telegram, Chatwork
-       |                            |
-       v                            v
- 3-Pass Extraction          Reactive Runtime Envelopes
-       |                    scope, expiry, signature, audit
-       +------------+---------------+
-                    |
-             MAMA Core (mama-memory.db)
-             memory, raw refs, model runs,
-             tool traces, twin edges, packets
-                    |
-             +------+------+
-             |             |
-        Viewer UI     Claude Code Plugin / MCP
-```
-
-## CLI
-
-| Command                                      | Description                          |
-| -------------------------------------------- | ------------------------------------ |
-| `mama init`                                  | Initialize workspace                 |
-| `mama setup`                                 | Print onboarding contract and status |
-| `mama start`                                 | Start daemon                         |
-| `mama stop`                                  | Stop daemon                          |
-| `mama status`                                | Show status and exact next actions   |
-| `mama gateway telegram --token-stdin`        | Validate and save a Telegram token   |
-| `mama gateway telegram detect-owner`         | Discover and confirm the owner chat  |
-| `mama report now`                            | Request and confirm the first report |
-| `mama connector <add\|remove\|list\|status>` | Manage connectors                    |
-
-## Configuration
-
-Main config: `~/.mama/config.yaml`
-
-| Variable                  | Default                  |
-| ------------------------- | ------------------------ |
-| `MAMA_DB_PATH`            | `~/.mama/mama-memory.db` |
-| `MAMA_HTTP_PORT`          | `3847`                   |
-| `MAMA_WORKSPACE`          | `~/.mama/workspace`      |
-| `MAMA_STAGE2_WORKORDERS`  | retired; unset or `on`   |
-| `MAMA_TEMPORAL_RECONCILE` | `off`                    |
-
-Timeout tuning lives under `timeouts` in `config.yaml`. The persistent CLI process pool supports:
-
-| Option                               | Default                     | Purpose                                     |
-| ------------------------------------ | --------------------------- | ------------------------------------------- |
-| `persistent_process_idle_ms`         | `session_ms`                | Reclaim idle backend sessions/processes     |
-| `persistent_process_cleanup_ms`      | `session_cleanup_ms`        | How often idle-process cleanup runs         |
-| `persistent_process_pending_tool_ms` | `max(4 * idle, 30 minutes)` | Max wait for pending tool-result handshakes |
-
-## Related Packages
-
-| Package                      | Purpose                             |
-| ---------------------------- | ----------------------------------- |
-| **@jungjaehoon/mama-os**     | Always-on AI runtime (this package) |
-| **@jungjaehoon/mama-server** | MCP server for Claude Desktop       |
-| **@jungjaehoon/mama-core**   | Shared memory engine                |
-
-## Development
-
-```bash
-git clone https://github.com/jungjaehoon-lifegamez/MAMA.git
-cd MAMA && pnpm install && pnpm build
-pnpm test       # 6,421 passing tests across all packages
-```
-
-## Links
-
-[GitHub](https://github.com/jungjaehoon-lifegamez/MAMA) · [npm](https://www.npmjs.com/package/@jungjaehoon/mama-os) · [Docs](https://github.com/jungjaehoon-lifegamez/MAMA/tree/main/docs) · [Issues](https://github.com/jungjaehoon-lifegamez/MAMA/issues)
-
-## License
-
-MIT
-
----
-
-**Last Updated:** 2026-08-28
+The guides use `mama` for the built CLI. Onboarding is terminal-only: the owner types tokens
+with echo off. It writes configuration and optional launchd files; backend login and service
+startup are separate steps.
+
+| Command                                       | Purpose                                          |
+| --------------------------------------------- | ------------------------------------------------ |
+| `mama init`                                   | Set up the owner, backend and connectors         |
+| `mama secret set <NAME>` / `mama secret list` | Rotate a token / list names only                 |
+| `mama daemon`                                 | Run the configured daemon in the foreground      |
+| `mama replay`                                 | Replay prepared historical inputs in day windows |
+| `mama status` / `mama stop`                   | Report running/stopped / stop the process        |
+
+See [CLI flags](../../docs/reference/cli.md) and
+[launchd management](../../docs/guides/troubleshooting.md).
+
+## Current surface
+
+- **Owner chat:** Telegram, with an allowed chat and owner sender.
+- **Five source connectors:** Chatwork, Slack, Trello, Kagemusha (read-only local bridge),
+  and Google Calendar through `gws`.
+- **Records:** tasks and revision history, a four-slot board, wiki pages, daily journals,
+  lessons, preferences and constraints.
+- **Reports:** live deltas, full reports at 08/13/18 KST and hourly reminders at 09–21.
+- **Viewer:** board, work, memory graph, wiki, logs and security events; read-only data routes.
+- **Files:** source attachment lookup/download, native workspace processing and
+  `deliver.telegram.file` to the configured owner.
+
+Assembly lives in `src/runtime/` and `src/cli/commands/daemon.ts`; actions in `src/api/`,
+collectors in `src/connectors/`, Telegram in `src/gateways/`, replay in `src/replay/`,
+and viewer sources/assets in `ui/` and `public/viewer/`.
+
+## Data, security and status
+
+OS state lives in `~/.mama/`; its database defaults to `~/.mama/memory.db`.
+`config.yaml` and `connectors.json` hold settings; terminal-entered credentials live in
+`auth.env` (0600). Both backends have workspace-only writes and credential-read exclusions.
+The viewer binds to `127.0.0.1:3847` by default; remote data access requires a bearer token
+or verified Access JWT. See [security](../../docs/guides/security.md).
+
+The [public MCP server](../mcp-server/README.md) and
+[plugin](../claude-code-plugin/README.md) run independently with a separate development database.
+
+[Owner loop](../../docs/explanation/owner-loop.md) · [Connectors](../../docs/guides/connectors.md) ·
+[Backends](../../docs/guides/backends.md) · [Reports](../../docs/guides/reports-and-board.md) ·
+[Replay](../../docs/guides/replay.md) · [Viewer](../../docs/guides/viewer.md).
+
+Completion is measured by [INTENT.md](../../INTENT.md); open live checks are in
+[checks.md](../../docs/rebuild/checks.md). Run `pnpm test` inside this package, with isolated
+state as described in [testing](../../docs/development/testing.md).
+
+[MIT](../../LICENSE).

package/db/migrations/001-connector-event-index.sql

@@ -0,0 +1,251 @@
+-- The connector event index, declared by the package that has connectors.
+--
+-- It was declared by the core, across seven core migrations, because a migration's
+-- identity was a bare version number and no other author could pick one. So every
+-- consumer of the core created these tables: the Claude plugin's live database
+-- holds eighteen of them and zero rows, and so does the MCP server's.
+--
+-- Three of the four consumers have no connectors. This is the one that does.
+--
+-- Declared whole rather than as a chain of seven, because it is being declared
+-- for the first time under this source. The statements are the schema those seven
+-- produced, read back from a built database rather than transcribed.
+
+CREATE TABLE connector_event_index (
+  event_index_id TEXT PRIMARY KEY,
+  source_connector TEXT NOT NULL,
+  source_type TEXT NOT NULL,
+  source_id TEXT NOT NULL,
+  source_locator TEXT,
+  channel TEXT,
+  author TEXT,
+  title TEXT,
+  content TEXT NOT NULL,
+  event_datetime INTEGER,
+  event_date TEXT,
+  source_timestamp_ms INTEGER NOT NULL,
+  metadata_json TEXT,
+  artifact_locator TEXT,
+  artifact_title TEXT,
+  content_hash BLOB NOT NULL CHECK(length(content_hash) = 32),
+  indexed_at TEXT NOT NULL,
+  updated_at TEXT NOT NULL,
+  expires_at TEXT
+, source_cursor TEXT, tenant_id TEXT, project_id TEXT, memory_scope_kind TEXT, memory_scope_id TEXT, operator_ingest_seq INTEGER CHECK (
+    operator_ingest_seq IS NULL OR operator_ingest_seq >= 1
+  ), operator_observation_seq INTEGER CHECK (
+    operator_observation_seq IS NULL OR operator_observation_seq >= 1
+  ), source_entity_id TEXT, current_observation_id TEXT REFERENCES observation_versions(observation_id));
+
+CREATE TABLE connector_event_index_cursors (
+  connector_name TEXT PRIMARY KEY,
+  last_seen_timestamp_ms INTEGER NOT NULL DEFAULT 0,
+  last_seen_source_id TEXT NOT NULL DEFAULT '',
+  last_sweep_at TEXT,
+  last_success_at TEXT,
+  last_error TEXT,
+  last_error_at TEXT,
+  indexed_count INTEGER NOT NULL DEFAULT 0
+);
+
+CREATE TABLE connector_event_index_observation_cursors (
+  source_connector TEXT PRIMARY KEY,
+  next_seq INTEGER NOT NULL CHECK (next_seq >= 1)
+);
+
+CREATE TABLE connector_event_index_operator_seq_cursors (
+  source_connector TEXT NOT NULL,
+  channel TEXT NOT NULL DEFAULT '',
+  next_seq INTEGER NOT NULL CHECK (next_seq >= 1),
+  PRIMARY KEY (source_connector, channel)
+);
+
+CREATE INDEX idx_connector_event_index_artifact_locator
+  ON connector_event_index(artifact_locator)
+  WHERE artifact_locator IS NOT NULL;
+
+CREATE INDEX idx_connector_event_index_cursor_order
+  ON connector_event_index(source_connector, source_timestamp_ms, source_id);
+
+CREATE INDEX idx_connector_event_index_event_date
+  ON connector_event_index(event_date, event_datetime, event_index_id);
+
+CREATE INDEX idx_connector_event_index_event_datetime
+  ON connector_event_index(event_datetime, event_index_id);
+
+CREATE INDEX idx_connector_event_index_expires_at
+  ON connector_event_index(expires_at)
+  WHERE expires_at IS NOT NULL;
+
+CREATE UNIQUE INDEX idx_connector_event_index_observation_seq
+  ON connector_event_index(source_connector, operator_observation_seq)
+  WHERE operator_observation_seq IS NOT NULL;
+
+CREATE INDEX idx_connector_event_index_operator_cursor_order
+  ON connector_event_index(source_connector, channel, operator_ingest_seq);
+
+CREATE UNIQUE INDEX idx_connector_event_index_operator_scope_seq
+  ON connector_event_index(source_connector, COALESCE(channel, ''), operator_ingest_seq)
+  WHERE operator_ingest_seq IS NOT NULL;
+
+CREATE UNIQUE INDEX idx_connector_event_index_source_identity
+  ON connector_event_index(source_connector, source_id);
+
+CREATE INDEX idx_connector_event_scope
+  ON connector_event_index(tenant_id, project_id, memory_scope_kind, memory_scope_id);
+
+CREATE INDEX idx_connector_event_source_cursor
+  ON connector_event_index(source_connector, source_cursor);
+
+CREATE INDEX idx_connector_event_source_entity
+  ON connector_event_index(source_connector, source_entity_id);
+
+CREATE TRIGGER trg_connector_event_index_legacy_content_refresh_au
+AFTER UPDATE OF content_hash, metadata_json, source_timestamp_ms, source_type, channel
+ON connector_event_index
+WHEN (
+  OLD.content_hash IS NOT NEW.content_hash
+  OR OLD.metadata_json IS NOT NEW.metadata_json
+  OR OLD.source_timestamp_ms IS NOT NEW.source_timestamp_ms
+  OR OLD.source_type IS NOT NEW.source_type
+  OR OLD.channel IS NOT NEW.channel
+) AND (
+  NEW.operator_ingest_seq IS OLD.operator_ingest_seq
+  OR NEW.operator_observation_seq IS OLD.operator_observation_seq
+)
+BEGIN
+  UPDATE connector_event_index
+  SET operator_ingest_seq = CASE
+        WHEN NEW.operator_ingest_seq IS OLD.operator_ingest_seq THEN NULL
+        ELSE NEW.operator_ingest_seq
+      END,
+      operator_observation_seq = CASE
+        WHEN NEW.operator_observation_seq IS OLD.operator_observation_seq THEN NULL
+        ELSE NEW.operator_observation_seq
+      END
+  WHERE event_index_id = NEW.event_index_id;
+END;
+
+CREATE TRIGGER trg_connector_event_index_observation_seq_ai
+AFTER INSERT ON connector_event_index
+WHEN NEW.operator_observation_seq IS NULL
+BEGIN
+  INSERT INTO connector_event_index_observation_cursors
+    (source_connector, next_seq)
+  VALUES (NEW.source_connector, 1)
+  ON CONFLICT(source_connector) DO NOTHING;
+  UPDATE connector_event_index
+  SET operator_observation_seq = (
+    SELECT next_seq FROM connector_event_index_observation_cursors
+    WHERE source_connector = NEW.source_connector
+  )
+  WHERE event_index_id = NEW.event_index_id;
+  UPDATE connector_event_index_observation_cursors
+  SET next_seq = next_seq + 1
+  WHERE source_connector = NEW.source_connector;
+END;
+
+CREATE TRIGGER trg_connector_event_index_observation_seq_au
+AFTER UPDATE OF operator_observation_seq ON connector_event_index
+WHEN NEW.operator_observation_seq IS NULL AND OLD.operator_observation_seq IS NOT NULL
+BEGIN
+  INSERT INTO connector_event_index_observation_cursors
+    (source_connector, next_seq)
+  VALUES (NEW.source_connector, 1)
+  ON CONFLICT(source_connector) DO NOTHING;
+  UPDATE connector_event_index
+  SET operator_observation_seq = (
+    SELECT next_seq FROM connector_event_index_observation_cursors
+    WHERE source_connector = NEW.source_connector
+  )
+  WHERE event_index_id = NEW.event_index_id;
+  UPDATE connector_event_index_observation_cursors
+  SET next_seq = next_seq + 1
+  WHERE source_connector = NEW.source_connector;
+END;
+
+CREATE TRIGGER trg_connector_event_index_observation_seq_explicit_ai
+AFTER INSERT ON connector_event_index
+WHEN NEW.operator_observation_seq IS NOT NULL
+BEGIN
+  INSERT INTO connector_event_index_observation_cursors
+    (source_connector, next_seq)
+  VALUES (NEW.source_connector, 1)
+  ON CONFLICT(source_connector) DO NOTHING;
+  UPDATE connector_event_index_observation_cursors
+  SET next_seq = CASE
+    WHEN next_seq <= NEW.operator_observation_seq THEN NEW.operator_observation_seq + 1
+    ELSE next_seq
+  END
+  WHERE source_connector = NEW.source_connector;
+END;
+
+CREATE TRIGGER trg_connector_event_index_operator_ingest_seq_ai
+AFTER INSERT ON connector_event_index
+WHEN NEW.operator_ingest_seq IS NULL
+BEGIN
+  INSERT OR IGNORE INTO connector_event_index_operator_seq_cursors (
+    source_connector,
+    channel,
+    next_seq
+  )
+  VALUES (NEW.source_connector, COALESCE(NEW.channel, ''), 1);
+
+UPDATE connector_event_index
+  SET operator_ingest_seq = (
+    SELECT next_seq
+    FROM connector_event_index_operator_seq_cursors
+    WHERE source_connector = NEW.source_connector
+      AND channel = COALESCE(NEW.channel, '')
+  )
+  WHERE event_index_id = NEW.event_index_id;
+
+UPDATE connector_event_index_operator_seq_cursors
+  SET next_seq = next_seq + 1
+  WHERE source_connector = NEW.source_connector
+    AND channel = COALESCE(NEW.channel, '');
+END;
+
+CREATE TRIGGER trg_connector_event_index_operator_ingest_seq_au
+AFTER UPDATE OF operator_ingest_seq ON connector_event_index
+WHEN NEW.operator_ingest_seq IS NULL AND OLD.operator_ingest_seq IS NOT NULL
+BEGIN
+  INSERT INTO connector_event_index_operator_seq_cursors
+    (source_connector, channel, next_seq)
+  VALUES (NEW.source_connector, COALESCE(NEW.channel, ''), 1)
+  ON CONFLICT(source_connector, channel) DO NOTHING;
+  UPDATE connector_event_index
+  SET operator_ingest_seq = (
+    SELECT next_seq FROM connector_event_index_operator_seq_cursors
+    WHERE source_connector = NEW.source_connector
+      AND channel = COALESCE(NEW.channel, '')
+  )
+  WHERE event_index_id = NEW.event_index_id;
+  UPDATE connector_event_index_operator_seq_cursors
+  SET next_seq = next_seq + 1
+  WHERE source_connector = NEW.source_connector
+    AND channel = COALESCE(NEW.channel, '');
+END;
+
+CREATE TRIGGER trg_connector_event_index_operator_ingest_seq_explicit_ai
+AFTER INSERT ON connector_event_index
+WHEN NEW.operator_ingest_seq IS NOT NULL
+BEGIN
+  INSERT OR IGNORE INTO connector_event_index_operator_seq_cursors (
+    source_connector,
+    channel,
+    next_seq
+  )
+  VALUES (NEW.source_connector, COALESCE(NEW.channel, ''), 1);
+
+UPDATE connector_event_index_operator_seq_cursors
+  SET next_seq = CASE
+    WHEN next_seq <= NEW.operator_ingest_seq THEN NEW.operator_ingest_seq + 1
+    ELSE next_seq
+  END
+  WHERE source_connector = NEW.source_connector
+    AND channel = COALESCE(NEW.channel, '');
+END;
+
+INSERT OR IGNORE INTO schema_version (source, version, description)
+VALUES ('standalone', 1, 'The connector event index belongs to the package with connectors');