- (databases) Use name not description for API alignment (#112)
- (datasets) Add missing
-o/--outputflag todatasets create; move success banner to stderr so-o jsonstdout isjq-parseable (#110) - (sandbox) Move "Sandbox created" and "Sandbox updated" banners to stderr for clean
-o jsonoutput (#110) - (sandbox) Fix missing trailing newline in
sandbox readoutput (#110) - (sandbox) Add
sandbox delete <id>subcommand; clears the active session automatically when the deleted sandbox was the active one (#110) - (workspaces) Fix incorrect lock check in
workspaces setβ was checkingHOTDATA_WORKSPACE(always set in sandbox runs), now correctly checksHOTDATA_SANDBOX(#110) - (context) Surface a friendly hint when
context pushis blocked inside an active sandbox, pointing users tohotdata sandbox set(no args) to clear it (#110)
- (skills) Bump skill file versions to 0.3.1 so
hotdata skills installcorrectly detects and installs the latest skills for CLI v0.3.x
- (query) Fetch results as Arrow IPC instead of JSON; reduces transfer size and preserves native types (#103)
- (query) Add
--database/-dflag to scope a query to a managed database without changing the active database (#102) - (databases) Add
databases show <id>as an explicit subcommand alias (#103) - (databases)
databases tables <id>now lists tables without requiring thelistsubcommand (#103) - (skills) Add
skills listas an alias forskills status(#103) - (update) Background update check with post-command notice; never blocks command output (#104)
- (update) Auto-install and update skills to match the new binary version during
hotdata update(#105) - (update) Execute
brew upgradedirectly for Homebrew installs instead of printing manual instructions (#106)
- (query) Async polling loop exits with code 2 on unexpected statuses instead of spinning forever (#103)
- (query) Failed async queries now surface the real server error message (#103)
- (query)
results get <id>now fetches Arrow IPC like the rest of the query path (#103) - (query) Polling loop polls first before checking the deadline, eliminating a mandatory 500ms delay (#106)
- (skills) Add 120-second HTTP timeout to the skills tarball download during
hotdata update(#106)
- (skills) Update skills to reflect recent API changes: database-scoped context,
databases set,--expires-at, corrected flag names fordatabases create/datasets create/datasets update(#100)
- (context) Scope context commands to active database;
hotdata contextnow calls/databases/{id}/contextand requires--database-idor an active database set viadatabases set(#98) - (databases) Add
--expires-atflag todatabases create; accepts relative durations (24h,7d) or RFC 3339 timestamps (#97) - (datasets) Remove upload/URL/file create paths;
datasets createnow requires exactly one of--sqlor--query-id(#95) - (databases) Migrate CLI to dedicated
/databasesAPI;databases setsaves active database;X-Database-Idheader sent automatically on all requests (#94)
- (datasets) Add missing
typediscriminator to dataset source payloads sent to API - (context) Correct
--database-idflag name in error message
- (indexes) Dot-bracket notation for
indexes create:airbnb.listings[col1,col2]replaces--connection-id/--schema/--table/--columns(#92) - (databases) Add
databases load <db.table>shorthand replacingdatabases tables load(#92) - (indexes) Make
--nameoptional onindexes create; auto-derived from table, columns, and type (#92)
- (databases) Remove
load:hint fromdatabases createsuccess output (#92)
- (search) Infer
--typeand--columnfrom table indexes; schema defaults topublic(#90)
- (search) Explicit error when a search index has no columns (#90)
- (databases) Add
--urlflag totables loadfor remote parquet files (#88)
- (auth) Add
hotdata auth registercommand (#85, #86) - (auth) Default register to GitHub; add
--emailflag - (update) Add
hotdata updatecommand - (skills) Split bundled skills into
hotdata-searchandhotdata-analytics(#84)
- (auth) Align CLI callback page colors with web app theme
- (auth) Extract
run_browser_authhelper; add tests forexchange_cli_register_code
- (skill) Epic flow checklists, datasets vs databases workflows, tag-only release finish (#84)
- (databases) Add managed databases CLI for parquet table loads (#82)
- (sandbox) Add sandbox JWT support
- (tty) Add no-input flag and tty checks for interactive commands
- (deps) Bump openssl to 0.10.79 for CVE fixes (#77)
- Ignore macOS metadata files (#81)
- (skill) Document managed databases commands
- (wizard) Render schema description, examples, defaults (#75)
- (skill) Align hotdata skill with CLI behavior
- (datasets) Add update subcommand to rename label or table_name
- Data/dataset refresh + indexes auto-embedding + embedding providers (#67)
- (skills) Add optional geospatial agent skill
- (skills) Auto-update bundled agent skills after CLI upgrade
- (datasets) Match runtimedb response shape on update
- (datasets) Drop synthetic schema_name on update output
- (datasets) Restore eprintln for "Dataset updated" status line
- (skills) Complete partial installs and improve status output
- (skills) Show Installed: No when no skill store exists
- (skills) Stop repeat auto-downloads (parse SKILL.md, stale tarball guard)
- (release) Bump geospatial skill version on release
- (skills) Always auto-update skills when eligible (remove env opt-out)
- (datasets) Drop redundant Stylize import in update path
- (auth) Add CLI auth session support (JWT access tokens, refresh, PKCE login)
- (indexes) Workspace-wide list with filters and parallel fetch
- (codecov) Treat patch coverage as informational
- Raise coverage for indexes list and get_none_if_not_found
- (auth) Add login subcommand mirroring bare auth
- (context) Strip .md suffix using correct byte length
- (context) Avoid UTF-8 panic when probing .md suffix
- (release) Pass --no-confirm to cargo release
- (skill) List before show; avoid blind context show DATAMODEL
- (skill) Context: notation and analysis vs DATAMODEL
- (context) Add context list/show/pull/push commands
- (context) Fail-fast pull when target exists; expand stem tests
- (release) Regenerate changelog with git-cliff
- (context) Clearer fetch_context exhaustiveness; drop no-op mkdir
- (cli) Clarify datasets command as upload and query Parquet/CSV
- (skill) Prefer workspace context API for data model and agents
- (skill) Context API only for data model and workspace docs
- (readme) Document workspace context commands and API-first model
- (skill) Align Hotdata SKILL with current CLI flags
- (skill) Sandbox datasets, long flags, and WORKFLOWS
- (skill) Unify dataset SQL as datasets..
- (sandbox) Align CLI with updated sandbox API
- (connections) Add health check to connection flow
- (sessions) Add session command feature
- (security) Update tar from 0.4.44 to 0.4.45
- (ci) Update workflow actions for Node.js 24 compatibility
- (ci) Revert release.yml (cargo-dist managed)
- (ci) SHA-pin codecov-action for consistency
- (ci) Keep create-github-app-token at v1 to avoid breaking changes
- (skill) Document results list, connections refresh, queries update flags
- (skill) Workflows, references, and SKILL alignment
- (auth) Fix config initialization and url upload api parameter
- (search) Add basic vector search (l2_distance)
- (query) Async query execution by returning early from long running queries
- (connections) Add connection information command (#28)
- (completions) Add completions command and include in brew formula
- (search) Add indexes and text search commands
- (queries) Add upload url for datasets and new queries commands (#23)
- (datasets) Add --sql and --query-id to datasets create
- (workspaces) Add set command to switch default workspace
- (results) Add results list command
- (connections) Add connections refresh endpoint
- (jobs) Add jobs commands
- (cli) Add table formatting and fix wrapping issues
- (cli) Skills/Auth CLI command quality of life improvements
- (skills) Add skill install command for CLI
- (datasets) Add datasets commands to create, list, and view datasets in CLI
- (datasets) Add upload id functionality to datasets create command
- (connections) Add connection command to list/view/create connections via the CLI
- (results) Add query result retrieval via results command
- (workspaces) Default workspace support for CLI commands
- (tables) Add search and pagination to tables list command
- Format table list output with full table path ..
- (tables) Add table list command and better table rendering
- (workspace) Add workspace list command
- (connection) Add connections list command
- (query) Add query execution command