# Ecosystem Implementation Tracker > **Status:** Active tracker > **Owner:** `learning_ai_common_plat` > **Purpose:** Single place to track cross-product ecosystem implementation progress, owners, status, and commit links. --- ## 1. Roadmap Summary ### Original Goal Build a shared ByteLyst ecosystem layer so products do not behave like isolated apps. The intended outcome was: - shared artifact, event, timeline, and runtime contracts in `learning_ai_common_plat` - real cross-product golden flows with preserved lineage - a first hosted timeline and runtime console - direct Cowork and FlowMonk runtime adoption - commit-linked progress tracking so implementation can be reviewed phase by phase ### What We Achieved - Phase 1 is complete: - LysnrAI transcript -> NoteLett note -> MindLyst memory - Phase 2 is complete: - FlowMonk plan -> ChronoMind routine -> EffoRise habit - Phase 3 is complete: - Cowork -> ActionTrail -> NoteLett -> MindLyst audited artifact flow - Phase 4 baseline is complete: - canonical timeline contract - hosted timeline ingest/query API - hosted internal timeline UI - Phase 5 baseline is complete: - canonical agent runtime contract - platform runtime projection API - hosted internal runtime console - direct Cowork runtime projections - direct FlowMonk runtime projections - direct FlowMonk runtime review UI - live Playwright verification of the FlowMonk runtime review path ### What Remains In The Roadmap - no remaining mandatory roadmap items - optional future product expansion may continue if real usage justifies it Canonical backlog: - [`OPTIONAL_PRODUCT_ENHANCEMENT_BACKLOG.md`](/Users/saravana/BytelystAI/learning_ai/learning_ai_common_plat/docs/ecosystem/OPTIONAL_PRODUCT_ENHANCEMENT_BACKLOG.md) ### Current Roadmap Call - the implementation roadmap baseline is materially complete through Phase 5 - the requested documentation hardening pass is complete - the currently tracked optional product enhancements are complete - any next work is net-new product expansion, not unfinished roadmap delivery ### Remaining Work Buckets #### Must-Finish Hardening - [x] expand adoption docs beyond the currently covered products and flows Owner: - `learning_ai_common_plat` Notes: - adoption coverage now includes SmartAuth, Auth App, mac-tooling, and JarvisJr in addition to the flow-driving repos - [x] harden the verification matrix with repo-level test ownership Owner: - `learning_ai_common_plat` Notes: - repo-level evidence owners and baseline command expectations are now documented - [x] revise the temporary ownership model if the working team changes Owner: - `learning_ai_common_plat` Notes: - revision triggers and next-state ownership shape are now documented while preserving the current four-party model - [x] define broader shared artifact storage and lifecycle rules Owner: - `learning_ai_common_plat` Notes: - storage class, retention, archival, migration, and binary attachment rules are now documented - [x] broaden the shared event taxonomy beyond the currently implemented phase seams Owner: - `learning_ai_common_plat` Notes: - additional canonical event families now cover marketplace, identity, delivery, sync, and timeline review domains - [x] harden approvals and trust policy details Owner: - `learning_ai_common_plat` Notes: - approval reuse, fresh-approval classes, degraded-device handling, and approval payload requirements are now documented - [x] finish marketplace inventory and entitlement details Owner: - `learning_ai_common_plat` Notes: - canonical listing schema, entitlement guidance, compatibility, and inventory mapping are now documented - [x] finish longer-term timeline hosting and filtering decisions Owner: - `learning_ai_common_plat` Notes: - timeline noise filtering and the current hosting choice are now documented explicitly #### Optional Product Enhancements - [x] evaluate richer FlowMonk replay execution controls Owner: - `learning_ai_flowmonk` Notes: - safe replay preview and guarded replay execution now exist for deterministic runtime actions - [x] evaluate richer Cowork checkpoint drill-in Owner: - `oss/learning_ai_claw-cowork` - `learning_ai_common_plat` Notes: - `cowork-service` now exposes one-checkpoint drill-in and the hosted admin runtime console now distinguishes direct checkpoint observations from projected summary fields Commits: - `82a44c2` --- ## 2. Tracking Rules Use this document as the execution ledger for ecosystem work. Rules: 1. Every implementation item should have exactly one status owner. 2. Every completed item should include at least one commit reference. 3. Use checkboxes for executable tasks and update them in the same PR/commit series that implements the work. 4. If a task spans multiple repos, list all affected repos in the row. 5. If a task is blocked, state the blocker explicitly instead of leaving it ambiguous. Status values: - `not-started` - `in-progress` - `blocked` - `done` Commit link format: - short hash only when local - GitHub commit URL once pushed Example: - `eae3409` - `https://github.com/saravanakumardb1/learning_ai_common_plat/commit/eae3409` --- ## 3. Core Spec Status | Spec | Status | Owner Repo | Notes | Commits | | ---------------------------------------------- | ----------- | ------------------------- | ---------------------------------------------------------------------------------------------------------------------------------- | --------------------------------------------------------------------------- | | `ECOSYSTEM_CROSS_POLLINATION_OPPORTUNITIES.md` | done | `learning_ai_common_plat` | strategy baseline created and strengthened | `d28e9d0`, `341cde7`, `3e572b6` | | `ECOSYSTEM_SHARED_ARTIFACT_SCHEMA.md` | done | `learning_ai_common_plat` | canonical artifact shape, lifecycle rules, migration guidance, and binary attachment handling are now documented | `eae3409`, `76f1b47`, `617d970` | | `ECOSYSTEM_AGENT_RUNTIME_CONTRACT.md` | in-progress | `learning_ai_common_plat` | lifecycle boundaries, queued runs, checkpoint/resume semantics, replay requirements, and direct Cowork/FlowMonk mappings now exist | `eae3409`, `ff8c5eb`, `97b731e`, `faf93ec`, `2ad979c`, `4bb8efe`, `3330ca5` | | `ECOSYSTEM_APPROVALS_AND_TRUST_MODEL.md` | done | `learning_ai_common_plat` | approval reuse, degraded-device handling, fresh-approval classes, and canonical approval payload requirements are now documented | `eae3409`, `066e75f` | | `ECOSYSTEM_EVENT_TAXONOMY.md` | done | `learning_ai_common_plat` | implemented phase seams plus broader event-family coverage are now documented | `41fa2cd`, `76f1b47`, `617d970` | | `ECOSYSTEM_OWNERSHIP_MATRIX.md` | done | `learning_ai_common_plat` | current four-party model, revision triggers, and next-state ownership shape are now documented | `41fa2cd`, `7a86a76`, `ab50831` | | `ECOSYSTEM_VERIFICATION_MATRIX.md` | done | `learning_ai_common_plat` | repo-level verification ownership and baseline evidence expectations are now documented | `41fa2cd`, `ab50831` | | `ECOSYSTEM_CONTRACT_TEST_STRATEGY.md` | in-progress | `learning_ai_common_plat` | executable fixture locations and runtime-route test anchors are now documented | `41fa2cd` | | `ECOSYSTEM_MARKETPLACE_UNIFICATION.md` | done | `learning_ai_common_plat` | listing schema, entitlement guidance, compatibility, and first inventory mapping are now documented | `7a86a76`, `066e75f` | | `ECOSYSTEM_PERSONAL_TIMELINE_PRD.md` | done | `learning_ai_common_plat` | timeline hosting choice, noise filtering, and long-term placement guidance are now documented | `7a86a76`, `066e75f` | --- ## 4. Phase 1 Checklist ### Foundation - [x] Create canonical ecosystem docs home in `docs/ecosystem/` Commits: - `5f4134e` - [x] Draft cross-pollination strategy Commits: - `d28e9d0` - `341cde7` - `3e572b6` - [x] Draft artifact schema stub Commits: - `eae3409` - [x] Draft runtime contract stub Commits: - `eae3409` - [x] Draft approvals and trust model stub Commits: - `eae3409` - [x] Draft event taxonomy Commits: - `41fa2cd` - [x] Implement Phase 1 artifact and event contracts in `@bytelyst/events` Commits: - `76f1b47` - [x] Draft temporary ownership matrix Commits: - `41fa2cd` - `7a86a76` - [x] Draft verification matrix Commits: - `41fa2cd` - [x] Draft contract-test strategy Commits: - `41fa2cd` - [x] Draft marketplace unification spec Commits: - `7a86a76` - [x] Draft personal timeline PRD Commits: - `7a86a76` - [x] Draft first adoption docs for transcript -> note -> memory flow Commits: - `7a86a76` ### First implementation tracks - [x] LysnrAI transcript -> NoteLett note -> MindLyst memory Repos: - `learning_voice_ai_agent` - `learning_ai_notes` - `learning_multimodal_memory_agents` - `learning_ai_common_plat` Commits: - `76f1b47` - `dd3d622` - `3b8298b` - `6ffc2f8` - `63c791f` - `79c85ad` - `9ecc7c6` - `0f469a9` - `49796b1` - `623d02c` - `dd3636d` Status note: - artifact flow is implemented end-to-end and downstream event lineage is now hardened with persisted latest-event indexes - MindLyst web Phase 1 memory-import path is verified by Codex with focused Vitest using a workspace `vitest` binary; `49796b1` improved the local package path but did not fully close it by itself - `learning_ai_notes` Phase 1 import path verified by Windsurf (`623d02c`): 13 focused tests, 3 schema conformance checks against `@bytelyst/events`, 106 total tests passing - Codex completed a manual runtime walkthrough on `2026-04-03` using `/tmp/bytelyst-phase1-walkthrough.qzzwch`, which produced transcript, note, and memory artifacts plus canonical event indexes with the causal chain `capture.transcript.created -> artifact.created -> artifact.linked -> memory.entry.created` - remaining known gap is local to `mindlyst-native/web`: package-local `npm run test:phase1` is still not self-sufficient in this clone even though the focused path is verified - [x] FlowMonk plan -> ChronoMind routine -> EffoRise habit Repos: - `learning_ai_flowmonk` - `learning_ai_clock` - `learning_ai_efforise` - `learning_ai_common_plat` Commits: - `78918fb` - `59ab8b1` - `fdd3743` - `4975498` - `f65f41c` Status note: - shared contract now supports first-class `plan`, `routine`, and `habit` artifacts - FlowMonk exports a canonical plan artifact plus latest plan-created event index - ChronoMind imports the latest plan into a routine artifact and preserves plan-created causation - EffoRise imports the latest routine into a habit artifact and preserves routine-linked causation - Codex completed a manual runtime walkthrough on `2026-04-03` using `/tmp/bytelyst-phase2-walkthrough.WpEjZu`, which produced the causal chain `evt_plan_created_b202f23f69a4 -> evt_routine_created_4c07d3fdcd07 -> evt_routine_linked_08b9db8a7013 -> evt_habit_created_f532f13332d1 -> evt_habit_linked_7be9a47928e3` - [x] Cowork -> ActionTrail -> NoteLett -> MindLyst audited artifact flow Repos: - `oss/learning_ai_claw-cowork` - `learning_ai_trails` - `learning_ai_notes` - `learning_multimodal_memory_agents` - `learning_ai_common_plat` Commits: - `e6b58b7` - `e294cf5` - `4af86b4` - `43819f1` - `a253de1` - `7ee2151` - `5a7f80d` Status note: - shared contract now supports a canonical `trail-report` payload while reusing generic `artifact.created`, `artifact.linked`, and `memory.entry.created` - ActionTrail imports Cowork audit-export JSON into a persisted `trail-report` artifact plus latest created-event index - NoteLett imports the latest `trail-report` into a markdown audit note and preserves upstream trail-report causation - MindLyst imports the latest audit note into an `insight` memory proposal and preserves upstream note-link lineage - Codex completed a manual runtime walkthrough on `2026-04-03` using `/tmp/bytelyst-phase3-walkthrough.myuu1z`, which produced the causal chain `evt_trail_created_b40197d5302e -> evt_note_created_abc1ad821974 -> evt_note_linked_9fd9a2b03363 -> evt_memory_8103b59fb40c` - review follow-up fixed the ActionTrail trace semantics so `causationId` and `parentEventId` no longer misuse Cowork task IDs - NoteLett now exposes `/api/ecosystem/phase3/import-latest-trail-report` - MindLyst now exposes the audited import path through `POST /api/memory` with `importLatestAuditNote: true` --- ## 5. Current Gaps Before Broad Implementation - [x] event taxonomy phase-1 event list hardened in `@bytelyst/events` - [x] adoption docs now cover the first flow plus SmartAuth, Auth App, mac-tooling, JarvisJr, Cowork, FlowMonk, ChronoMind, EffoRise, NoteLett, ActionTrail, and MindLyst - [x] ownership model now documents how it should change if the team expands beyond the current four-party operating model - [x] verification matrix now includes repo-level test ownership and baseline evidence expectations - [x] contract-test strategy fixture location and validation tooling now exist in `packages/events` These are now resolved for the current documentation baseline. --- ## 6. Phase 4 Checklist - [x] define canonical `TimelineItem` contract in `@bytelyst/events` Commits: - `3f2482b` - [x] define baseline event-to-timeline inclusion rules for Phase 1 to Phase 3 Commits: - `3f2482b` - [x] implement pure timeline aggregation over canonical ecosystem events Commits: - `3f2482b` - [x] verify that Phase 1 to Phase 3 events render in one unified stream Commits: - `3f2482b` - [x] add first hosted timeline ingest/query API in `platform-service` Commits: - `e377351` Status note: - platform-service now persists timeline items in `timeline_items` - `POST /api/timeline/ingest` accepts canonical ecosystem events and stores unified timeline items - `GET /api/timeline` returns the unified timeline stream with non-admin JWT callers scoped to their own data - [x] add first hosted internal timeline UI in `admin-web` Commits: - `6f3a563` - `d328c7a` Status note: - admin-web now exposes `/timeline` over the shared platform timeline API - the first hosted internal timeline UI supports user, artifact, correlation, and event-name filtering - the timeline hardening pass adds explicit product-scope review, correlation-chain grouping, and focused helper tests --- ## 7. Phase 5 Checklist - [x] define canonical runtime schemas for dispatch, session, task, todo, run, approval, and action logs Commits: - `3f2482b` - [x] validate Cowork-style remote dispatch against the shared runtime contract Commits: - `3f2482b` - [x] validate FlowMonk-style scheduled execution against the shared runtime contract Commits: - `3f2482b` - [x] add first runtime projection + dispatch validation API in `platform-service` Commits: - `fe36296` Status note: - `GET /api/agent-runtime/sessions` projects platform sessions into the shared `AgentSession` shape - `GET /api/agent-runtime/runs` projects platform runs into the shared `AgentRun` shape - `POST /api/agent-runtime/dispatch/validate` validates `AgentDispatchRequest` payloads against the canonical schema - [x] add first hosted internal runtime console in `admin-web` Commits: - `71ef2ac` Status note: - admin-web now exposes `/agent-runtime` over the shared platform runtime API - the first hosted internal runtime UI supports projected session review, projected run review, and dispatch payload validation - [x] wire first product implementations to emit the shared runtime objects directly from Cowork and FlowMonk Status note: - Cowork product-native runtime projections are now implemented in cowork-service - `023826e` adds `GET /api/agent-runtime/sessions`, `GET /api/agent-runtime/runs`, `GET /api/agent-runtime/approvals`, and `POST /api/agent-runtime/dispatch/validate` - `01201f8` adds `GET /api/agent-runtime/tasks` with canonical `AgentTask` projection - `b8242b4` adds `GET /api/agent-runtime/actions` with canonical `AgentActionLog` projection - `97b731e` adds `GET /api/agent-runtime/todos` with an interim task-backed `AgentTodo` projection - FlowMonk local installs now resolve `@bytelyst/*` from the sibling `learning_ai_common_plat` workspace instead of the dead localhost registry - `1ccafa7` adds FlowMonk direct runtime projections for sessions, tasks, runs, action logs, and dispatch validation - `faf93ec` adds FlowMonk direct approval and todo projections - [x] promote `queued` to a first-class `AgentRun` state and preserve it in shared runtime projections Commits: - `ff8c5eb` Status note: - `AgentRun` now supports `queued` directly in `@bytelyst/events` - platform-service now preserves queued platform runs as `queued` - cowork-service now projects pending IPC tasks as queued runs - FlowMonk now projects scheduled entries as queued runs - [x] emit and consume canonical Cowork audit event IDs across Rust, cowork-service, and ActionTrail Commits: - `3963f63` - `a57b367` - `37e9a3c` Status note: - Cowork Rust now emits canonical `event_id` values on audit records and forwards them into platform audit `details.eventId` - cowork-service now prefers the canonical event ID for `AgentApprovalCheckpoint.approvalId` and `AgentActionLog.actionLogId` - ActionTrail now preserves the latest Cowork audit `event_id` as the `trail-report` provenance origin action and `artifact.created` causation anchor ### 6.1 Remaining Direct Runtime TODOs - None for the Phase 5 baseline. Remaining work is optional product enhancement, not contract completion. ### 6.2 Explicit Blockers And Questions - No blocking Phase 5 questions remain. - Optional product question: should FlowMonk grow from runtime review into richer replay execution controls later? --- ## 7. Active Execution Plan Current implementation kickoff document: - [`PHASE1_TRANSCRIPT_NOTE_MEMORY_EXECUTION_PLAN.md`](/Users/saravana/BytelystAI/learning_ai/learning_ai_common_plat/docs/ecosystem/PHASE1_TRANSCRIPT_NOTE_MEMORY_EXECUTION_PLAN.md) - [`PHASE2_PLAN_ROUTINE_HABIT_EXECUTION_PLAN.md`](/Users/saravana/BytelystAI/learning_ai/learning_ai_common_plat/docs/ecosystem/PHASE2_PLAN_ROUTINE_HABIT_EXECUTION_PLAN.md) - [`PHASE3_COWORK_ACTIONTRAIL_NOTE_MEMORY_EXECUTION_PLAN.md`](/Users/saravana/BytelystAI/learning_ai/learning_ai_common_plat/docs/ecosystem/PHASE3_COWORK_ACTIONTRAIL_NOTE_MEMORY_EXECUTION_PLAN.md) - [`PHASE4_PERSONAL_TIMELINE_EXECUTION_PLAN.md`](/Users/saravana/BytelystAI/learning_ai/learning_ai_common_plat/docs/ecosystem/PHASE4_PERSONAL_TIMELINE_EXECUTION_PLAN.md) - [`PHASE5_AGENT_RUNTIME_CONTRACT_EXECUTION_PLAN.md`](/Users/saravana/BytelystAI/learning_ai/learning_ai_common_plat/docs/ecosystem/PHASE5_AGENT_RUNTIME_CONTRACT_EXECUTION_PLAN.md) Execution status: - shared contract baseline implemented in [`packages/events`](/Users/saravana/BytelystAI/learning_ai/learning_ai_common_plat/packages/events) - LysnrAI emits canonical transcript artifact + `capture.transcript.created` - LysnrAI also persists the latest transcript event index for downstream causation (`79c85ad`) - NoteLett imports the latest transcript artifact into a note, emits `artifact.created` + `artifact.linked`, and now preserves upstream capture event lineage (`9ecc7c6`) - MindLyst imports the latest note artifact into a memory proposal, emits `memory.entry.created`, and now preserves upstream note-link lineage (`0f469a9`) - MindLyst focused tests were verified by Codex with `/Users/saravana/BytelystAI/learning_ai/learning_ai_common_plat/node_modules/.bin/vitest run src/lib/ecosystem-phase1.test.ts src/__tests__/api-memory.test.ts`; `49796b1` improved the local script/setup path - Codex completed an end-to-end runtime walkthrough on `2026-04-03` rooted at `/tmp/bytelyst-phase1-walkthrough.qzzwch`; the persisted indexes show `evt_capture_e913dd681aba -> evt_note_created_cced1442727e -> evt_note_linked_c3d477dfd0c4 -> evt_memory_9a3ebd9c67ca` - repo-local verification for MindLyst web Phase 1 path is functionally confirmed, but the package-local `npm run test:phase1` path is still not self-sufficient in this clone - Windsurf verified the `learning_ai_notes` Phase 1 backend path with `cd backend && npx vitest run`; 22 files passed and 106 tests passed, including the focused ecosystem-phase1 coverage - shared contract now also supports the Phase 2 `plan -> routine -> habit` seam through `artifact.created` and `artifact.linked` - Phase 2 walkthrough is complete and tracked in [`PHASE2_PLAN_ROUTINE_HABIT_EXECUTION_PLAN.md`](/Users/saravana/BytelystAI/learning_ai/learning_ai_common_plat/docs/ecosystem/PHASE2_PLAN_ROUTINE_HABIT_EXECUTION_PLAN.md) - Phase 3 walkthrough is complete and tracked in [`PHASE3_COWORK_ACTIONTRAIL_NOTE_MEMORY_EXECUTION_PLAN.md`](/Users/saravana/BytelystAI/learning_ai/learning_ai_common_plat/docs/ecosystem/PHASE3_COWORK_ACTIONTRAIL_NOTE_MEMORY_EXECUTION_PLAN.md) - Phase 4 timeline baseline is complete and tracked in [`PHASE4_PERSONAL_TIMELINE_EXECUTION_PLAN.md`](/Users/saravana/BytelystAI/learning_ai/learning_ai_common_plat/docs/ecosystem/PHASE4_PERSONAL_TIMELINE_EXECUTION_PLAN.md) - Phase 5 runtime-contract baseline is complete and tracked in [`PHASE5_AGENT_RUNTIME_CONTRACT_EXECUTION_PLAN.md`](/Users/saravana/BytelystAI/learning_ai/learning_ai_common_plat/docs/ecosystem/PHASE5_AGENT_RUNTIME_CONTRACT_EXECUTION_PLAN.md) - FlowMonk runtime review is now live-e2e verified with Playwright, and runtime todo/approval mutations emit replayable source-domain events (`71038bb`, `5fc778b`)