All notable changes to Supertag CLI are documented in this file.
The format is based on Keep a Changelog, and this project adheres to Semantic Versioning.
Export Status Display Bug - Fixed supertag-export status always showing "Token expired" even when logged in
Export Token Refresh for Automated Exports - Fixed automated/scheduled exports failing after 60 minutes
supertag-export login once after updating to enable automated exportsCore Node Fields in Aggregate --where - WHERE clause now supports filtering by core node fields like name
--where "name~keyword" returned 0 resultssupertag aggregate --tag todo --group-by Priority --where "name~urgent"Default Output Format - Changed default output from JSON to table (TSV)
--format json or --json for JSON outputAggregate Table Alignment - Fixed column alignment for entries containing emojis
Query Field Output with Select Clause - Include custom field values in query output
select * returns all supertag fields including inherited fieldsselect "Email,Phone" returns specific fields by nameis empty operator for querying empty/missing field valuessupertag query "find contact select *"supertag query "find task where Status is empty"Semantic Search --min-score Option - Filter results by minimum similarity threshold
supertag search "query" --semantic --min-score 0.5 - Only results with >= 50% similarity@Name Reference Resolution - Reference nodes by display name instead of ID
@Name syntax in field values: --state "@Open"--assignees "@Alice,@Bob"supertag create task "My Task" --state "@Open"embed and server commands that depend on native modulesField Default Values - Auto-populate field default values when creating nodes
Unified Field Format - Both MCP and CLI now accept fields in two formats
{ "fields": { "Status": "Done" } }{ "Status": "Done" }Interactive Launchd Setup in Installer - Installer now prompts to set up background services during installation
--no-launchd flag to skip this stepInteractive Menu for install-launchd.sh - Running without parameters now shows a menu instead of defaulting to server
Scripts Not Included in Install - Fixed install-launchd.sh and other scripts not being available after installation
~/.supertag-cli/scripts/ for launchd setup and maintenanceInteractive Prompts in curl|bash Install - Fixed install prompts not working when running via curl | bash
NODE_PATH Unbound Variable - Fixed "NODE_PATH: unbound variable" error when NODE_PATH not previously set
supertag-export failing with "Cannot find package 'playwright'" error immediately after installationexport NODE_PATH in install script for current sessionsupertag tags list showing node count as 1 for all supertagssupertag-mcp binary crashing on startup with Claude DesktopDeleted Supertags Not Removed on Sync - Fixed issue where deleted supertags remained in the database after re-syncing
supertag sync index now properly removes supertags that no longer exist in your Tana exportWindows LanceDB Loading - Fixed startup crash on Windows when LanceDB native module wasn't available
sync index no longer require LanceDB to be installedsupertag attachments list command - List all attachments with metadata (filename, extension, tags)supertag attachments extract command - Download attachments to local directorysupertag attachments get <nodeId> command - Download a single attachment by node IDsupertag attachments stats command - Show attachment statistics by type and tagflat (default), date, tag, node--skip-existing flag-e png -e jpg or --extension pdf-t meeting or --tag projectsupertag update install failing with "operation not supported on socket, copyfile" errorDocuments\Tana-Export like macOS~/.local/share/supertag/exports (XDG-style) which didn't match the documented locationrelated command and tana_related MCP toolsupertag related <nodeId> - find all related nodessupertag related <nodeId> --direction in - incoming references onlysupertag related <nodeId> --direction out - outgoing references onlysupertag related <nodeId> --types child,reference - filter by relationship typesupertag related <nodeId> --depth 2 - multi-hop traversal (up to 5)tana_related { nodeId: "abc123", direction: "both", depth: 2 }Field References in Graph Traversal - The related command now correctly finds field references
Search Tag Query Filter - The search command now respects the query when combined with --tag
supertag search "Velo" --tag topic now returns only #topic nodes whose name contains "Velo"tana_tagged tool now supports optional query parameter for the same filteringSystem Field Discovery - Automatic detection and inclusion of Tana system fields
tags fields --all with system: true flagtana_supertag_info tool returns system field information[system] marker for system fields in table outputContains Operator Shorthand - Query filters now support ~value shorthand for contains queries
where: { name: "~meeting" } is equivalent to where: { name: { contains: "meeting" } }\~value to search for literal tilde prefixComma-Separated Reference Fields - Reference field values can now be comma-separated strings
"abc123, def456" are automatically split into arraysaggregate command and tana_aggregate MCP toolsupertag aggregate --tag task --group-by Statussupertag aggregate --tag meeting --group-by monthtana_aggregate { find: "task", groupBy: ["Status"] }--show-percent for percentages, --top N for top groupsWindows Native Build - Windows binaries now build on native Windows runner instead of cross-compilation
--semantic) and embedding commands now work on WindowsDocumentation - Removed outdated tana-daily script references
docs/export.md with current supertag-export and supertag sync commandsLAUNCHD-SETUP.md for macOS automationfindAllTagsByName now queries supertag_metadata instead of supertag_fieldstags show command_ownerId for "TRASH" markerisNodeInTrash() helper function for recursive ownership checkTrash Detection for Supertags - Supertags in Tana's trash are now filtered from queries
_ownerId property containing "TRASH"Playwright Export Build - Fixed supertag-export binary failing with "Cannot find package 'playwright'"
--external flags for playwright and related packagesbun add -g playwrightNODE_PATH environment variable pointing to global node_modulesInheritance-Aware Tag Search - New --include-descendants flag for search --tag command
search --tag "Source | Origin" --include-descendants returns nodes tagged with todo, project, Area, meeting, etc. (40 descendant tags)tags visualize commandSchema Registry Export (Spec 081) - Enhanced schema-registry.json with target supertag metadata
FieldSchema interface now includes optional targetSupertag property with id and nameUnifiedSchemaService.toSchemaRegistryJSON() exports target supertag for reference fieldsTarget Supertag Extraction (Spec 077) - Reference fields now store actual target supertag from Tana definition
target_supertag_id and target_supertag_name in supertag_fields tabletags fields command now includes targetSupertagId and targetSupertagName in JSON outputisTagId() now correctly distinguishes Tana IDs from kebab-case tag names/^[A-Za-z0-9_-]{8,}$//[A-Z]/ and /[a-z]/ patterns for ID detection2024-12-08) now work for created/updated fields7d, today) were converted to timestampscreated > 2024-12-07 and created < 2024-12-09Unified Query Language (Spec 063) - SQL-like query syntax for complex filtering
supertag query "find task where Status = Done" CLI commandtana_query MCP tool with structured inputfind <tag> where <conditions> order by <field> limit <n>= (exact), ~ (contains), >, <, >=, <=, exists, notand, or with parentheses groupingtoday, yesterday, 7d, 1w, 1m, 1yparent.tags, parent.nameselect parameter-field for descending orderBatch Operations (Spec 062) - Fetch or create multiple nodes in a single request
supertag batch get <ids...> command to fetch multiple nodes by IDecho "id1\nid2" | supertag batch get --stdinsupertag batch create command to create multiple nodes at oncesupertag batch create --file nodes.jsoncat nodes.json | supertag batch create --stdin--dry-runtana_batch_get and tana_batch_createUniversal Format Options (Spec 060) - Extended output formatting from 3 modes to 6 formats
--format <type> option on all standard commandsjson, table, csv, ids, minimal, jsonltable format for terminals, json for pipesSUPERTAG_FORMAT environment variable for default formatoutput.format setting--json and --pretty flagsCsvFormatter: RFC 4180 compliant CSV with proper escapingIdsFormatter: One ID per line for xargs pipingMinimalFormatter: JSON projection to id, name, tags onlyJsonlFormatter: JSON Lines format for streaming--no-header option for CSV/table formatsENOTSUP: operation not supported on socket error when installing updatesinstallUpdate function now properly unlinks the target file before copyingsupertag update check command to check for available updatessupertag update download command to download updates to local cachesupertag update install command with automatic backup and rollback on failureupdateCheck: 'enabled' | 'disabled' | 'manual' in configUnified Logger (Spec 057) - Migrated core infrastructure to centralized logging system
Batch Workspace Processor (Spec 056) - New processWorkspaces() utility for multi-workspace operations
Query Builder Utilities (Spec 055) - Safe SQL query construction helpers
buildPagination() for LIMIT/OFFSET with safe defaultsbuildOrderBy() with column whitelist validationbuildWhereClause() for parameterized WHERE conditionsbuildSelectQuery() for complete query compositionOutput Formatter Consolidation (Spec 054) - Unified output formatting across CLI
OutputFormatter interface with Unix, Pretty, and JSON implementationsresolveOutputMode() helper for consistent flag handlingcreateFormatter() factory for formatter instantiationDatabase Resource Management (Spec 053) - RAII-style database connection handling
withDatabase() for auto-closing database connectionswithQueryEngine() for combined database + query engine handlingwithTransaction() for automatic commit/rollbackwithWorkspaceDatabase() and withWorkspaceQuery() for workspace-resolved pathsDatabaseNotFoundError for better error messages--children JSON with nested structures silently discarded the nested childrenparseChildren() only extracted name, id, dataType - ignored children propertyparseChildObject() helper to properly handle nested structures--children '{"name": "Section", "children": [{"name": "Item"}]}' works correctlytana_create MCP tool and createNode() now support hierarchical child structureschildren arrays for deep nesting (e.g., workshop notes with sections and sub-points)z.lazy() for type-safe nested structuressupertag create) and MCP (tana_create) support nested children{"name": "Section 1", "children": [{"name": "Point 1"}, {"name": "Point 2"}]}Code Generation from Supertags (Spec 024) - Generate type-safe Effect Schema classes from Tana supertag definitions
supertag codegen generate -o <path> command for generating TypeScript code.extend()option (default), undefined, nullable--split generates separate files per supertag with barrel index--dry-run previews output without writing files--tags TodoItem Meeting generates only specified supertags--no-metadata)TUI Todo Example Application (Spec 031) - Terminal-based todo manager demonstrating codegen feature
examples/tui-todo/ with full documentationCodegen Documentation - New comprehensive documentation at docs/codegen.md
Explicit Field Types in Node Creation - createNode() now uses database field types for Input API payloads
{ dataType: "date", name: "2025-01-15" }{ dataType: "reference", id: "nodeId" }{ dataType: "url", name: "https://..." }supertag create) and MCP (tana_create) commandsDuplicate Supertag Handling - Fixed UnifiedSchemaService.getSupertag() selecting wrong entry
tags fields now shows field IDs and types matching tags show format- FieldName (fieldId) with Type: dataType on next line- FieldName (fieldId, from ParentTag)formatFieldLines() helper ensures DRY, consistent output across commandstags show - New --all flag shows inherited fieldssupertag tags show todo --all - Shows all 8 fields including inheritedtags fields --allSupertagMetadataService.getAllFields() for accurate inheritance chainSystem Field Markers in SchemaRegistry - tags show now displays system fields (Due Date, Date, etc.)
tags show showed fewer fields than tags fieldsSYSTEM_FIELD_MARKERS mapping from supertag-metadata.ts (DRY)Explicit Field Type Extraction - Field types now extracted from Tana's typeChoice structure
typeChoice tuples with SYS_D* codes:email, options for more accurate field typingValue-Based Type Inference - Fallback inference from actual field values
_metaNodeId (e.g., Horizon, Assignee)Transcript Search Table Format - transcript search --pretty now uses table format matching documentation
CI Test Fixes - Fixed MCP transcript tests failing on GitHub CI
Transcript Filtering and Commands (Spec 023) - Dedicated transcript access with default exclusion from search/embeddings
supertag transcript list command - List meetings with transcripts (ID, name, line count, date)supertag transcript show <id> command - Display transcript content with speaker and timing infosupertag transcript search <query> command - Full-text search within transcript content only--include-transcripts flag for embed generate to opt-in to transcript embeddingtana_transcript_list, tana_transcript_show, tana_transcript_search--pretty flag for transcript list commandExpanded System Field Mappings - Added 12 additional SYS_* field mappings for comprehensive field extraction
Transcript Search Meeting Context - Fixed tana_transcript_search returning null meetingId/meetingName
parent_id column (reduced from 13s to ~4s)_ownerId ending with _TRASH)Inline Reference Parsing - Meeting names with inline date references now display properly
<span data-inlineref-date=...> to readable ISO-8601 datesSystem Field Extraction - Fixed SYS_* fields (Due date, Date, Attendees) not being extracted into field_values table
SYS_A61 are synthetic IDs that don't exist in the nodes tableisFieldTuple() now recognizes SYS_* first children as valid field labelsresolveFieldNameFromTuple() now looks up SYS_* names from centralized SYSTEM_FIELD_NAMES mappingRe-embedding recommended for transcript functionality:
If you want semantic search to include transcript content, regenerate your embeddings:
# Default: transcripts excluded (recommended for most users)
supertag embed generate
# Include transcripts in semantic search (90K+ additional nodes)
supertag embed generate --include-transcripts
Transcript-specific search is always available via supertag transcript search regardless of embedding settings
3D Visualization Enhancements
Ancestor Traversal - New --from <tag> option for visualizing inheritance upwards
--root <tag> which shows descendantsSchema Consolidation (Spec 020) - Unified database-backed schema storage
UnifiedSchemaService class for database-backed schema queriesschema-registry.json cache after sync index (no separate schema sync needed)normalized_name, description, inferred_data_type for fieldssupertag_metadata table for tag-level metadata (name, description, color)getSchemaRegistryFromDatabase() function for loading schema from databasegetTagDetailsFromDatabase() function with inferred data typesbuildNodePayloadFromDatabase() function for node creation from databasetana_supertag_info tool now returns inferredDataType in field infonormalizeName(), inferDataType()System Field Extraction - Added support for Tana's built-in system fields
SYS_A90 → "Date" field (used by meeting, calendar-item, appointment supertags)SYS_A61 → "Due Date" field (used by task, todo, project supertags)Mp2A7_2PQw → "Attendees" field (used by meeting supertag)supertag tags fields <tag> outputSYSTEM_FIELD_MARKERS for extensibilitySupertag Visualization - New tags visualize command for inheritance graph visualization
--root <tag> for subtrees, --orphans to include isolated tags--direction (BT/TB/LR/RL), --show-fields, --show-inherited, --colors, --theme--output <file>, auto-open with --opendocs/visualization.md with rendering instructionsInteractive HTML Visualization - Self-contained HTML files with UML-style class diagram nodes
--theme option3D Visualization (Spec 021) - Interactive 3D graph visualization using Three.js
--format 3d generates self-contained HTML with 3d-force-graph--layout hierarchical mode positions parents above children--size-by-usage scales node size by tag usage count--theme dark/light for color scheme preferenceField Details in All Formats - --show-fields now shows actual field names in all visualization formats
--show-inherited displays inherited fields with their origin tagDuplicate Supertag Detection - Fixed inconsistent tag resolution when multiple tags share the same name
tags show, tags fields, and tags inheritance now use consistent preference logic (most inheritance parents, then most fields)supertag tags show fbAkgDqs3k)Tags Show Command - Fixed "Invalid export format" error in supertag tags show <tag>
getSchemaRegistry() which correctly parses cached registry formatSupertag Parent Extraction - Fixed extraction of parent supertag relationships
supertag_parents tablesupertag tags inheritance <tag> now shows complete inheritance treeRelease Packages Include Documentation - Release zips now include:
docs/ - Full documentation (MCP setup, embeddings, webhooks, launchd, etc.)launchd/ - macOS LaunchAgent plist templatesscripts/ - Installation and management scriptsTana Command Integration Guide - New documentation for setting up Tana Commands
${sys:context}, format, workspace)Sync Scheduler Runs Every 6 Hours - Daily sync now runs 4 times per day
supertag sync indexImproved launchd Documentation - Path configuration section added
ch.invisible.tana-* to ch.invisible.supertag-*supertag sync index commandThis is a major version release with significant CLI restructuring. Legacy commands have been removed.
Removed Commands (use new equivalents):
supertag query search → use supertag searchsupertag query tagged → use supertag search --tagsupertag query stats → use supertag stats --dbsupertag query top-tags → use supertag tags topsupertag query refs → use supertag nodes refssupertag query recent → use supertag nodes recentsupertag show node → use supertag nodes showsupertag show tagged → use supertag search --tag --showsupertag embed search → use supertag search --semanticsupertag embed stats → use supertag stats --embedsupertag embed filter-stats → use supertag stats --filterNew commands following the object action pattern for consistency and discoverability:
supertag search <query> - Unified search command
--semantic flag for vector similarity search--tag <name> flag for filtering by supertag--show flag for full node content display--depth <n> for child traversal with --showsupertag nodes show|refs|recent - Node operations
nodes show <id> - Display node contents with depth traversalnodes refs <id> - Show references to a nodenodes recent - Recently updated nodessupertag tags list|top|show - Supertag operations
tags list - List all supertagstags top - Most used supertagstags show <name> - Show tag schemasupertag stats - Unified statistics
--db - Database statistics only--embed - Embedding statistics only--filter - Content filter breakdownNew RESTful endpoints with consistent API design:
POST /search - Unified search with type parameter
type=fts - Full-text search (default)type=semantic - Vector similarity searchtype=tagged - Search by supertagGET /stats - Unified statistics with type parameter
type=all - All statistics (default)type=db - Database stats onlytype=embed - Embedding stats onlytype=filter - Content filter statsRESTful /nodes endpoints
GET /nodes/:id - Get node by ID with optional depthGET /nodes/:id/refs - Get node referencesGET /nodes/recent - Recently created nodesPOST /nodes/find - Find nodes by pattern/tagRESTful /tags endpoints
GET /tags - List all supertagsGET /tags/top - Top supertags by usageGET /tags/:name - Get tag schema detailsDeprecated webhook endpoints (still functional, marked for removal):
POST /semantic-search → use POST /search with type=semanticGET /embed-stats → use GET /stats?type=embedPOST /refs → use GET /nodes/:id/refsPOST /nodes → use POST /nodes/findPOST /tags → use GET /tags or GET /tags/topentry.value.apiKey from Firebase IndexedDBconfig.json instead of .envsupertag-export --login saves Firebase API key to ~/.config/supertag/config.jsonTANA_FIREBASE_API_KEY environment variable if not in config.env files with FIREBASE_API_TOKEN can be safely deletedAutomatic Browser Installation - supertag-export now auto-installs Chromium browser on first run
bunx playwright install chromium neededsupertag-export setup command for explicit installationAutomatic Workspace Configuration - supertag-export discover now auto-adds first workspace as main
supertag workspace add step for primary workspaceexport/supertag-export to ./supertag-exportmcp/supertag-mcp to ./supertag-mcpbun install from root instead of cd export && bun installFirebase Token Extraction - supertag-export --login now automatically extracts Firebase API token
.env as FIREBASE_API_TOKEN.env or creates new oneOpen Source Release - Repository is now publicly available on GitHub
github:jcfischer/resona#main)supertag activate, supertag deactivate, supertag license status removed~/.local/share/supertag/license.json) are no longer used but harmless if presentsupertag embed generate to create new LanceDB embeddings~/.config/supertag/config.json instead of databasesupertag embed generate to regenerate embeddings in LanceDB formattana-daily now generates embeddings after sync--embed flag to run embeddings only--no-embed flag to skip embeddings--all flagSymlink Resolution - tana-daily now properly resolves symlinks on macOS
~/bin/ and still find export toolsSync Command - Fixed sync to use main CLI with proper workspace support
--all mode now correctly indexes all enabled workspaces"Meeting with <span data-inlineref-node=\"id\">John</span> today""Meeting with <span data-inlineref-node=\"id\">John</span>"Inline Reference Support - Create inline references using Input API syntax
<span data-inlineref-node="NODE_ID">Display Text</span> in node names or childrenCLI Children Option - New -c, --children flag for supertag create
--children "Child text"--children '{"name": "Link", "id": "abc123"}'--children "See <span data-inlineref-node=\"xyz\">Related</span>"MCP Tool Documentation - tana_create now documents correct reference formats
<span data-inlineref-node="ID">Text</span>{"name": "...", "id": "..."}Embedding Cleanup - Sync now cleans up embeddings for deleted/modified nodes
sqlite-vec Fix - Indexer now loads custom SQLite before database creation
# Create todo with children and inline reference
supertag create todo "Meeting with <span data-inlineref-node=\"abc123\">John</span>" \
--children "First subtask" \
--children '{"name": "Assigned to", "id": "person123"}'
/workspaces endpoint lists available workspacesworkspace parameter to target specific workspace/health returns workspace list and default workspace infoStart server (serves all configured workspaces):
supertag server start --daemon
Query specific workspace:
# GET endpoints: use query param
curl "http://localhost:3000/stats?workspace=work"
# POST endpoints: include in body
curl -X POST http://localhost:3000/search \
-H "Content-Type: application/json" \
-d '{"query": "meeting", "workspace": "personal"}'
setEmbeddingConfig() only checked for dimension changes, not first-time setupmxbai-embed-large (1024d) now work correctly on fresh workspaces--fix flag to embed config command for manual repair of affected databasesIf you configured embeddings before v0.11.1 and got dimension mismatch errors:
supertag embed config --fix -w <workspace>
supertag embed generate -w <workspace>
embed config - Configure embedding provider and modelembed generate - Generate embeddings for nodes with content filteringembed search - Semantic search with similarity scoresembed stats - Show embedding statisticsprops._flags)isEntity(), isEntityById(), findNearestEntityAncestor() functionstana_semantic_search MCP tool for AI assistantswithDbRetry() and withDbRetrySync() for lock handlingbun run test - Fast tests onlybun run test:slow - Slow integration testsbun run test:full - Complete test suitebun run precommit - Alias for full suitesupertag-export discover now uses Tana's appState.nodeSpace.openFiles instead of network traffic capturesizeBytes from discovered workspace data (not available in appState)isRootFile flag to identify user's primary workspacepage.evaluate() to query Tana's internal state directly in browser contextpackage.jsonsrc/version.ts module exports VERSION constantpackage.jsonsupertag-export now works when called via symlink from any directoryBASH_SOURCE[0] returned symlink path instead of actual script locationsupertag-export command now works from any working directorysearch-filter.ts) used by CLI, MCP, and webhook server[[Something]] literal names) from search resultsembed filter-stats now shows entity detection statistics at the end_flags or inferred detectionembed stats now displays the database path being queriedminLength from 15 to 3 characters*, ..) still filteredprops._flags (with underscore prefix) from Tana exportstana-dump.ts now includes _flags and _entityOverride fields.passthrough() to PropsSchema to preserve additional underscore-prefixed props during parsing_flags with underscore, not flags. Zod was stripping unknown fields._flags=1 (previously only ~12k via tag/library inference)LENGTH(name) >= minLength OR is_entity ensures meaningful short entities get embeddedmigrateEmbeddingSchema() now checks if embeddings table exists before ALTERgetEmbeddingConfig() now checks if embedding_config table exists before queryingembed stats and embed filter-stats now work gracefully without embeddings configured- Search Results: Pizza %%view:table%% (colon confused Tana parser)- Search Results %%view:table%% (matches semantic search format)[[Name^nodeID]] syntax instead of creating duplicate nodes_TRASH ancestors to reduce deleted node pollutionnomic-embed-text to mxbai-embed-largesupertag embed generate)ancestor_id and context_text columns for traceabilitybuildContextualizedNode() function uses findMeaningfulAncestor() for context resolutionsupertag embed generateengine.rawDb instead of engine.db for ancestor resolution-w, --workspace <alias> flagembed generate --all-workspaces processes all enabled workspaces sequentiallyembed generate--min-length increased from 10 to 15 characters--min-length 10 to restore previous behavior if neededBuffer.from(embedding.buffer) without byteOffset/byteLengthBuffer.from(embedding.buffer, embedding.byteOffset, embedding.byteLength)supertag embed generatetana_semantic_search - New MCP tool for vector similarity searchminSimilarity threshold filteringincludeContents parameter for full node details (fields, children, tags)depth parameter for child traversal (0-3) when includeContents is truequery search, tana_search) and semantic search (embed search, tana_semantic_search)--no-ancestor (CLI) or includeAncestor: false (MCP)supertag embed filter-stats - New command showing filtering breakdown by docType--min-length <n> - Set minimum name length (default: 10)--include-all - Bypass all content filters--include-timestamps - Include timestamp-like nodes--include-system - Include system docTypes-v, --verbose - Show detailed filter information--show flag displays full node contents (fields, children, tags) in search results--depth <n> enables child traversal when using --show (default: 0)-a, --ancestor / --no-ancestor - Enable/disable ancestor resolution (default: enabled)show node command for consistent outputembed generate now applies smart content filtering by defaultembed stats now shows content filter statisticstana_semantic_search)query search now shows ancestor context by default (use --no-ancestor to disable)tana_search MCP tool now includes includeAncestor parameter (default: true)getNodeContents, formatNodeOutput, and related functions from show.ts for reusefindMeaningfulAncestor from ancestor-resolution.ts for reuserawDb getter to TanaQueryEngine for direct SQLite accesssupertag embed config - Configure embedding provider and modelsupertag embed generate - Generate embeddings for indexed nodessupertag embed search <query> - Semantic similarity searchsupertag embed stats - Show embedding statistics and coveragevec0.dylib/vec0.so) placed alongside the binary--quiet flag to mcphost examples for cleaner output--created-after, --created-before, --updated-after, --updated-beforesearch, nodes, tagged, and recent commandstana_search and tana_tagged also support date range filteringsupertag-mcp binary providing Model Context Protocol servertana_search, tana_tagged, tana_stats, tana_supertags, tana_node, tana_create, tana_synctana_create supports creating nodes with field values and child referencestana_sync enables triggering reindex or checking sync status from AI toolschildren parameter to tana_create for proper reference/link creation{name, id} create clickable links to existing nodes[[text^nodeId]] syntax doesn't work in node namesstoreData object~/.config/tana/ to ~/.config/supertag/release.sh script for automated builds and releases--push option for automatic git tag pushingsupertag sync cleanup command to remove old export files--keep N option--dry-run mode to preview deletionssupertag show node <id> -d <depth> for traversing child nodes--json flagtana to supertag for claritych.invisible.supertag-dailysupertag activate <key>supertag license status to check license statesupertag deactivate to free up activation slotssupertag (main CLI) + supertag-export (browser automation)supertag-export runsupertag workspace add <id> --alias <name> for managing multiple workspacessupertag workspace list and supertag workspace set-default--all flag (export, sync, cleanup)-w <alias>supertag-export discover --addsupertag CLIsupertag-export login for initial authenticationsupertag-export run for automated exportssupertag query search <term> with relevance rankingsupertag query tagged <supertag> for filtering by tagsupertag query stats for database statisticssupertag query top-tags for supertag usage countssupertag create <supertag> <name> for node creation--field value syntaxsupertag postsupertag formatsupertag server start with daemon modeSupertag CLI is developed by InVisible GmbH, Switzerland.