# Remove dead SQLite dual-path code; keep migration-only readers ## Summary PostgreSQL cutover left hundreds of production dual-path branches (`backendMode ? PG : SQLite/store.db`) whose SQLite arms only hit throwing `Database`/`ArchiveDatabase`/`CentralDatabase` stubs. This change mechanically collapses those unreachable arms so production authority is AsyncDataLayer/PostgreSQL only, while preserving the six authorized read-only migration/recovery `DatabaseSync` seams. ## Dual-path mass removed | Metric | Before | After | |---|---|---| | `if (…backendMode)` (non-test) | ~328 | ~70 | | `store.db` / `this.db` refs in core (non-test) | ~570+ | ~375 (mostly pure legacy MissionStore/eval/insight SQLite classes + thin getters) | | Net diff | — | **~6.7k lines removed** across 41 files | Remaining `backendMode` checks are intentional (incomplete-PG sync safe-defaults, settings-sync disabled-on-PG, symbol-lock PG-only gates, “requires PostgreSQL” config versioning throws), not live SQLite authority. ## Subsystems cleaned - **Core TaskStore / task-store/***: collapsed if/else and early-return dual-path across reads, moves, lifecycle, mutations, workflow, archive, branch/PR, artifacts, comments, audit, project ops, etc. `initImpl` is PostgreSQL-only (SQLite startup tail deleted). - **Satellite stores**: automation, agent, routine, plugin, secrets, approval-request, central-core dual-path arms collapsed. - **Plugins**: reports async methods, compound-engineering pipeline + session stores, CLI Printing Press store — SQLite fallbacks removed; PG required. - **Engine**: no functional dual-path change beyond whitespace (settings-sync / peer-exchange PG-disabled behavior kept). ## Six migration-only readers retained (allowlist unchanged) 1. `packages/core/src/postgres/sqlite-migrator.ts` 2. `packages/core/src/project-identity.ts` 3. `packages/core/src/sqlite-validation.ts` 4. `packages/core/src/postgres/startup-factory.ts` 5. `packages/cli/src/commands/db.ts` 6. `scripts/lib/start-local-project.mjs` Plus low-level `sqlite-adapter` and migrator/startup-import tests. Inventory ratchet still requires exactly these six `new DatabaseSync(` production sites, all `readOnly: true`. ## Not treated as SQLite - `.fusion/project.json`, `task.json`, `agent-log.jsonl` file storage - AsyncDataLayer / Drizzle PG paths - Incomplete-PG sync safe-default stubs (still return empty/false/null under backend without consulting SQLite) ## Verification - `sqlite-production-reader-inventory.test.ts` — 15/15 pass - `incomplete-pg-ports.pg.test.ts` — 6/6 pass - Targeted PG tests (create-task, move, handoff, runtime-persistence, agent, mission, insight, central-core) — green - `tsc --noEmit` for `@fusion/core`, `@fusion/engine`, `@fusion/dashboard` — green - `scripts/check-no-getdatabase.mjs` — clean <!-- This is an auto-generated comment: release notes by coderabbit.ai --> ## Summary by CodeRabbit * **Improvements** * Improved end-to-end consistency by making PostgreSQL/async persistence the standard across core task/workflow, automation, agents, plugins, routines, secrets, approvals, central operations, and session storage. * Unified scheduling, settings, configuration revision writes, run/workflow selection, queues/leases/transitions, and audit/lifecycle updates around consistent async transaction behavior. * **Bug Fixes** * Fixed edge cases for archived/deleted reads, unarchive/recovery flows, not-found handling, and task/artifact/document/log/comment operations, including more reliable emissions and hydration across search/list and lifecycle operations. <!-- end of auto-generated comment: release notes by coderabbit.ai -->
Reports Plugin for Fusion
Generates HTML system activity reports with multi-agent review.
Install (one-click)
- Open Settings → Plugins → Fusion Plugins.
- In Bundled Plugins, click Install on Reports.
- Enable the plugin if prompted.
Once installed and enabled, Fusion registers the Reports dashboard destination automatically.
Scaffold seams (interim)
The plugin currently exports four interim scaffold seams to unblock downstream implementation work:
resolveEnabledCadences/ReportsCadence(src/cadence.ts) — interim cadence-resolution seam; scheduled cadence registry + cron/sentinel wiring lands in FN-3779.aggregateReportData+ aggregation types (src/aggregation.ts) — interim aggregation seam; real aggregation orchestration lands in FN-3780.startReportsPipeline+ pipeline dependency interfaces (src/pipeline.ts) — interim orchestrator seam to keep call sites stable while FN-3779/FN-3780 wire real runtime components.createInMemoryReportsRunsStore+ run record/store types (src/runs-store.ts) — interim in-memory run state store; persistent store replacement lands in FN-3784.
Review Panel
The plugin exposes runReviewPanel() / runGeneratedReportReview() to fan out a generated report draft to multiple reviewer agents in parallel.
Panel member settings shape
Each reviewer uses this contract:
{
id: string;
name: string;
perspective: string;
promptTemplateId?: string;
provider?: string;
modelId?: string;
}
perspectiveis appended to the reviewer system prompt.promptTemplateIdselects a template fromsettings.reviewPromptTemplates[templateId]when present.provider+modelIdoptionally override model selection per reviewer.
Prompt template contract
runReviewPanel resolves reviewer templates in this order:
settings.reviewPromptTemplates[promptTemplateId ?? id]settings.reviewPrompt- Built-in fallback (
DEFAULT_REVIEW_PROMPT)
This is the temporary compatibility contract until FN-3782 lands shared review-template helpers.
Individual review shape
{
memberId: string;
memberName: string;
perspective: string;
verdict: "approve" | "revise" | "reject";
summary: string;
highlights: string[];
lowlights: string[];
suggestions: string[];
rawText: string;
durationMs: number;
}
Combined review shape
{
overallVerdict: "approve" | "revise" | "reject";
consensusSummary: string;
mergedHighlights: string[];
mergedLowlights: string[];
mergedSuggestions: string[];
individual: IndividualReview[];
failures: ReviewFailure[];
}
Aggregation is deterministic:
- verdict precedence:
approve < revise < reject - merged arrays are case-insensitive de-duped, first-seen order, max 25 items each
- consensus summary is generated locally from reviewer summaries (no second AI call)
Timeout and failure semantics
- Each reviewer has a hard timeout (
120_000ms). - A single reviewer failure never aborts the full panel.
- Failures are returned as:
{
memberId: string;
reason: "timeout" | "parse_error" | "session_unavailable" | "exception";
message: string;
}
- If all reviewers fail, combined verdict is
rejectwith an explicit consensus summary describing panel failure.
Report Archive
The plugin persists generated reports in the project PostgreSQL schema through the shared ReportStore provider. Rows are scoped by canonical project identity.
Schema
Table: reports
- identity/metadata:
id,cadence,title,metadataJson - period window:
periodStart,periodEnd - lifecycle/status:
status,failureReason - payload references:
draftMarkdown,renderedHtmlPath - review payload:
combinedReviewJson - timestamps:
generationStartedAt,generationCompletedAt,reviewStartedAt,reviewCompletedAt,approvedAt,publishedAt,archivedAt,createdAt,updatedAt - approval actor:
approvedBy
Indexes:
idxReportsCadenceCreatedon(cadence, createdAt DESC, id)idxReportsStatusUpdatedon(status, updatedAt DESC, id)idxReportsPeriodon(periodStart, periodEnd, id)
Status lifecycle
generating → review_pending → review_in_progress → review_complete → approved → published
failed and archived are allowed from any non-terminal state. Idempotent transitions (from === to) are no-ops.
Approval + publish lifecycle (FN-3787)
A parallel approvalState gate now controls human/approver decisions before distribution:
review_complete entry:
approvalRequired=false, autoPublishOnApproval=false→approvalState=approved,status=approvedapprovalRequired=false, autoPublishOnApproval=true→approvalState=published,status=publishedapprovalRequired=true→approvalState=awaiting_approval,status=review_complete
Decision transitions:
awaiting_approval --approve--> approved(or directlypublishedwhenautoPublishOnApproval=true)awaiting_approval --reject--> rejectedapproved --publish--> published
Backfilled legacy rows use approvalState=not_required and are non-actionable.
Authorization rules:
- When
approvalRequired=trueandapproverAgentIdsis non-empty, only listed approver agent IDs may approve/reject/publish. - When
approvalRequired=trueandapproverAgentIds=[], any human dashboard user is allowed; agents are not. publishTargetsrecords publish intent metadata when a report reachespublished.
Share-ready summary blocks (FN-3787)
Approved/published reports can produce deterministic share artifacts via GET /reports/:id/share-blocks:
plainText: compact paste-ready summarymarkdown: heading/bullets + report linkslack: mrkdwn-friendly summaryemailHtml: inline-styled HTML snippet for email clients
share-blocks is intentionally locked (409) until approvalState is approved or published.
Email HTML styling exemption:
emailHtmldeliberately uses inline style attributes and hardcoded hex colors for email-client compatibility; dashboard design-token CSS rules do not apply to this serialized output format.
ReportStore API
createReport(input)getReport(id)listReports(filter?)updateReport(id, patch)setStatus(id, next, opts?)attachReview(id, combinedReview)attachRenderedHtml(id, htmlPath)deleteReport(id)
Emitted events:
report:createdreport:updatedreport:status-changedreport:review-attachedreport:deleted
This archive is the source of truth for downstream report HTML rendering (FN-3785) and dashboard report list/detail flows (FN-3786).
Dashboard view
The plugin registers a primary dashboard view (Reports) via dashboardViews with componentPath: "./dashboard-view".
The view provides:
- History list of reports with filters (cadence, status, period date range, title search, agent filter)
- Embedded detail preview using sandboxed iframe + preview HTML endpoint
- Section quick-jump navigation by stable
data-sectionmarkers - Side-by-side comparison drawer for two reports with section-level diff summary
- Standalone HTML download action wired to the export endpoint