Repository navigation
aitools: emit Claude Code's native plugin install hint - #7585
Conversation
When Claude Code runs the CLI without the Databricks plugin or skills, the install hint now also prints Claude Code's native `<claude-code-hint>` line (https://code.claude.com/docs/en/plugins/cli-hints), which makes Claude Code offer a one-click install of `databricks@claude-plugins-official`. - The native line prints only when CLAUDE_CODE_CHILD_SESSION is set, outside CI, and when no standard stream is a terminal. It shares the text hint's per-session throttle and logs a CLAUDE_CODE_NATIVE hint event. - Only tooling Claude Code itself loads (Agent.DatabricksTooling) silences the hint; skills only in the shared ~/.databricks stores no longer do. - `auth token`, `auth docker token` and CLI processes started by the CLI (DATABRICKS_CLI_SELF_EXEC, set by `air images push` and `apps init`) print no install hint. - `cmdio.HasTerminal` (any standard stream is a TTY) and `cmdio.FakeTTYReader` for tests. - Telemetry docs: one `aitools_hint_event` per printed hint line. Co-authored-by: Isaac <no-reply@databricks.com>
028eb23 to
07a831d
Compare
|
Commit: Regression test reportSelected 3 changed tests (limit=20, 0 not selected) 3 tests: regression=3 unreleased=0 coverage=0 failing=0
TestAccept/air/image-push: branch ✅ | main (e1316e5) ❌ | latest (v1.26.0) ❌main (e1316e5): latest (v1.26.0): TestAccept/cmd/aitools/claude-config-dir: branch ✅ | main (e1316e5) ❌ | latest (v1.26.0) ❌main (e1316e5): latest (v1.26.0): same output as main (e1316e5) TestAccept/cmd/aitools/telemetry/hint: branch ✅ | main (e1316e5) ❌ | latest (v1.26.0) ❌main (e1316e5): latest (v1.26.0): |
✅ Integration test reportCommit: a057fdd
Top 4 slowest tests (at least 2 minutes):
|
simonfaltum
left a comment
There was a problem hiding this comment.
🤖 Automated review from simonfaltum's review agent (pr-review skill) at 07a831d. Each finding was verified against the code; simonfaltum may follow up.
No blocking issues found, approving. 10 non-blocking comment(s) and 3 question(s) below.
This PR prints Claude Code's native <claude-code-hint> line next to the existing text hint, bases Claude Code install detection on what Claude Code itself loads, and stops hinting in auth token, auth docker token and CLI processes started by the CLI. The tag format, stderr placement and CLAUDE_CODE_CHILD_SESSION gate match Claude Code's docs, the tests are thorough, and CI is green. I found nothing blocking. There are two nice-to-have points: a subagent or hook call can use up the session's throttle without ever showing the dialog, and apps init still treats shared-store skills as installed for Claude Code. There are also a few nits and three non-blocking questions about scope, the single-agent abstraction and the extra CI and terminal gates.
Questions
- This PR changes four things: the native hint; which installs count for Claude Code (shared-store skills no longer count, and project or local plugins count only inside their project); the
auth token/auth docker tokenand self-exec exclusions; and theagentHintrefactor. AGENTS.md asks for one change per PR. The exclusions look like a prerequisite for the native line, because Claude Code can't strip a tag embedded mid-line. Would you move the detection change, and possibly the refactor, into separate PRs, or do they need to land together? (non-blocking) agentHint,hintForAgent(a switch with one case), the nil check innativeLine, the empty check insessionIDandTestAgentHint_NativeLineNilIsEmptyexist for agents that don't have a hint yet. AGENTS.md asks for simplicity and no speculative flexibility. Would you keep the direct Claude Code check plus a few lines for the native line, and add the table when a second agent needs it? (non-blocking)- The cli-hints docs only ask for the
CLAUDECODE/CLAUDE_CODE_CHILD_SESSIONgate, and Claude Code itself never prompts in-p, subagent, hook or unattended runs. The CI and terminal checks add an 11-variable CI list to three acceptance scripts. If some Claude Code surface runs tools with a terminal on any stream, the native line silently never prints, and telemetry would not show it because it only counts printed lines. Was the dialog checked end to end in an attended session (the description notes Windows and macOS were not tested)? Are both checks needed? (non-blocking)
Comments outside the changed lines
- [nit]
libs/aitools/agents/hint_test.go:254: This test still writessome-other-skillinto the shared store, whichclaudeToolingInstalledno longer reads, so it passes whatever the prefix filter does. UsewriteClaudeSkillso it checks that a non-Databricks skill in~/.claude/skillsdoesn't silence the hint. - [nit]
.github/PULL_REQUEST_TEMPLATE.md: The description uses## Summary,## Evidenceand## Merge Danger..agents/skills/pr-checklist/SKILL.mdasks for the template's## Changes,## Whyand## Tests, in that order and with no extra sections, and recent merged PRs follow it. Suggest moving the content under those headings.
| if native := hint.nativeLine(ctx); native != "" { | ||
| cmdio.LogString(ctx, native) | ||
| logHintShown(ctx, hint.nativeKind) | ||
| } | ||
| cmdio.LogString(ctx, hint.text) | ||
| logHintShown(ctx, hint.textKind) | ||
| recordSessionHint(ctx, hint.sessionEnv) |
There was a problem hiding this comment.
[nice to have] The session marker is written whenever the text line prints, including when the native line printed somewhere Claude Code never shows the dialog. A subagent's Bash calls carry the parent's CLAUDE_CODE_SESSION_ID and CLAUDE_CODE_CHILD_SESSION=1, and the cli-hints docs say subagent runs and hook output strip the tag without prompting. So when a subagent or a hook makes the session's first databricks call, the main loop's commands print neither line for the next hour, and a session shorter than an hour never sees the dialog. The same happens when the native line is withheld because a terminal is attached. Consider throttling the native line under its own key, and not writing that key when the line was withheld. A shorter TTL for the native line would also limit the subagent and hook case, since Claude Code already shows the dialog at most once per session.
There was a problem hiding this comment.
Good catch. Addressed in #7692 (stacked on this one): the native line now has its own per-session marker with a 10-minute TTL, separate from the text hint's hour, and it is written only when the native line actually prints, so a withheld line (terminal attached, CI, no child-session marker) no longer burns it. The CLI can't tell a subagent or hook from the main loop, so the short TTL is what bounds that case; Claude Code shows the dialog at most once per session, so repeating the line is cheap.
| // claudeToolingInstalled reports whether Claude Code has Databricks AI tooling | ||
| // in its own locations (see Agent.DatabricksTooling): the databricks plugin, or | ||
| // databricks skills in Claude's skills dirs. Skills present only in the shared | ||
| // ~/.databricks/aitools/skills store don't count, because Claude Code doesn't | ||
| // load them from there. Any read/parse failure reports false, so an unreadable | ||
| // state errs toward hinting. |
There was a problem hiding this comment.
[nice to have] This comment now says Claude Code doesn't load skills from the shared ~/.databricks/aitools/skills store. But HasDatabricksToolingInstalled (skills.go:23-33), which apps init uses, still treats shared-store skills as covering every agent, Claude Code included. With skills only in the shared store, databricks apps init under Claude Code prints this hint ("Databricks skills are not installed…") from the root hook, while the command's own check reports the tooling as installed and skips its tip. A human running apps init with Claude Code detected is never offered the install either. Either switch HasDatabricksToolingInstalled to per-agent DatabricksTooling, or say in both comments why they differ. A follow-up PR is fine.
There was a problem hiding this comment.
Agreed, the two shouldn't disagree. I'll move HasDatabricksToolingInstalled to per-agent DatabricksTooling in a separate follow-up PR to keep this one scoped.
| tooling := claude.DatabricksTooling(ctx, func() bool { return false }) | ||
| if tooling == ToolingPlugin { | ||
| version, _ := claude.DatabricksPluginVersion(ctx) | ||
| log.Debugf(ctx, "Databricks plugin detected for Claude Code (version %q); skipping install hint", version) |
There was a problem hiding this comment.
[nit] WithCallingAgentToolingInUserAgent (cmd/root/root.go:88) already computes DatabricksTooling for Claude Code right before MaybeHint. On the plugin path, DatabricksPluginVersion then parses installed_plugins.json a third time just for a debug log, and because the throttle is never written once the plugin is installed, this runs on every command. Also, func() bool { return false } duplicates noRecordedPlugin in recommend.go:45, and the Go style guide asks for a named function. Suggest dropping the version from the debug line and adding a package-level noRecordedPlugin for both callers.
There was a problem hiding this comment.
Done in a057fdd: a single package-level noRecordedPlugin in calling.go, shared with recommend.go, and the debug line no longer reads the plugin version, so installed_plugins.json isn't parsed again on every command.
| export CLAUDE_CODE_CHILD_SESSION=1 | ||
| unset CLAUDE_CODE_SESSION_ID | ||
| # CI and the CI/CD provider markers the SDK detects (GitHub Actions sets both). | ||
| unset CI GITHUB_ACTIONS GITLAB_CI JENKINS_URL TF_BUILD CIRCLECI TRAVIS BITBUCKET_BUILD_NUMBER PROJECT_ID CODEBUILD_BUILD_ARN TFC_RUN_ID |
There was a problem hiding this comment.
[nit] The same unset CI GITHUB_ACTIONS … TFC_RUN_ID line, which mirrors the SDK's CI provider list, appears here, in telemetry/hint/script:7 and in air/image-push/script:69, and again as hintEnvVars in hint_test.go. If the SDK adds a provider, the native line disappears only when tests run on that provider's CI, and four copies need updating. .agents/rules/testing.md says shared bash helpers go in a parent script.prepare, and an unset_ci_env helper there would keep a single copy.
There was a problem hiding this comment.
Moved into one unset_ci_env helper in the root acceptance/script.prepare in #7691 (the only parent the three scripts share; it only defines a function, so other tests are unaffected). Its comment points at the SDK's listCiCdProviders so the list stays in sync. Goldens are unchanged, and the three tests pass with CI=true GITHUB_ACTIONS=true exported.
| // in parallel (see testutil.SetAgentEnv). The first t.TempDir call runs before | ||
| // SetAgentEnv clears TMPDIR / TEMP, so later calls in the test still work. | ||
| // It also moves into an empty project with its own .git boundary, so project | ||
| // skills in the checkout running the tests are not detected. | ||
| func setClaudeAgent(t *testing.T) { | ||
| cwd := t.TempDir() | ||
| require.NoError(t, os.Mkdir(filepath.Join(cwd, ".git"), 0o755)) | ||
| t.Chdir(cwd) | ||
| testutil.SetAgentEnv(t, "CLAUDECODE", "1") | ||
| unsetEnv(t, hintEnvVars) |
There was a problem hiding this comment.
[nit] testutil.SetAgentEnv already clears the environment (NullEnvironment → os.Clearenv) and keeps TMPDIR, TEMP and TMP. So unsetEnv(t, hintEnvVars) right after it, here and at line 467, does nothing, and the sentence saying SetAgentEnv clears TMPDIR/TEMP is wrong. Suggest dropping those two calls and that sentence. hintEnvVars is then needed only in TestClaudeToolingInstalled_LogsPluginVersion, where only CLAUDE_CONFIG_DIR matters.
There was a problem hiding this comment.
You're right, those calls did nothing. Removed in a057fdd together with the wrong sentence; hintEnvVars and unsetEnv had no callers left, so they're gone too. The plugin-log test now just calls testutil.SetAgentEnv(t).
| // Claude Code's own manifest decides, so the CLI's install record is never | ||
| // consulted for it. | ||
| tooling := claude.DatabricksTooling(ctx, func() bool { return false }) |
There was a problem hiding this comment.
[nit] Plugin records with project or local scope now count only when the CLI's working directory is inside the recorded project. Claude Code loads them from the directory it started in, which the CLI can't see. So a session started in ~/repo with a project-scoped plugin that runs cd /tmp/scratch && databricks jobs list is now told "Databricks skills are not installed". Before this PR, any plugin record kept the hint quiet. The description and the changelog mention the shared-store change but not this one. Please add it there, or keep counting project and local records if a false hint is worse than a missed one.
There was a problem hiding this comment.
Keeping the behaviour, since it matches what Claude Code actually loads for that directory, but it's now documented in the claudeToolingInstalled comment, the changelog fragment and the PR description, with the cd /tmp/x && databricks ... example (a057fdd).
| // TF_BUILD, which doesn't set CI). | ||
| func inCI(ctx context.Context) bool { | ||
| if v := env.Get(ctx, ciEnv); v != "" { | ||
| if b, err := strconv.ParseBool(v); err != nil || b { |
There was a problem hiding this comment.
[nit] strconv.ParseBool fails on no and off, so CI=no or CI=off counts as CI here and the native line is dropped. libs/versioncheck reads the same variable with env.GetBool, which treats those values as false. Treat no and off as false too, by checking them before the ParseBool fallback.
There was a problem hiding this comment.
Fixed in a057fdd, though not via env.GetBool: it returns false for values it can't parse, so CI=woodpecker (providers that set CI to their own name) would stop counting as CI. inCI now treats the same false values GetBool accepts ("", 0, f, false, no, off, case-insensitive) as not CI and anything else as CI; a comment explains why, and tests cover CI=no and CI=OFF.
| @@ -0,0 +1 @@ | |||
| * When Claude Code runs the CLI without the Databricks plugin or skills, the CLI now also prints Claude Code's native `<claude-code-hint>` line, at most once per hour per Claude Code session together with the text hint, so Claude Code offers a one-click install of the Databricks plugin. Skills stored only in `~/.databricks/aitools/skills` or the legacy `~/.databricks/agent-skills` no longer count as installed for Claude Code, and `databricks auth token`, `databricks auth docker token` and the CLI commands that `databricks air images push` and `databricks apps init` run as subprocesses no longer print install hints. ([#7585](https://github.com/databricks/cli/pull/7585)) | |||
There was a problem hiding this comment.
[nit] .agents/skills/pr-checklist/SKILL.md asks for entries that start with a verb, or with the affected command followed by "now". This one opens with a "When…" clause. At 678 characters it is also the longest fragment in .nextchanges/cli and puts four changes in one line. Suggest leading with the change, e.g. "Print Claude Code's native plugin install hint when…", and keeping the detection and auth token notes short.
There was a problem hiding this comment.
Rewritten in a057fdd to lead with "Print Claude Code's native plugin install hint…", cut from 678 to 498 characters, and it now mentions the project/local plugin scope change.
- inCI: treat CI=no/off as not CI, matching env.GetBool's false values, while any other non-empty value (e.g. a provider name) still means CI. - claudeToolingInstalled: share a package-level noRecordedPlugin with MaybeVersionCheckTip and drop the plugin version from the debug log, which re-parsed installed_plugins.json on every command. - Document that project/local plugin records count only when the CLI's cwd is inside the recorded project. - Tests: check a non-Databricks skill in ~/.claude/skills, drop the no-op env unsets after testutil.SetAgentEnv, cover CI=no/OFF. - Shorten the changelog fragment and lead with a verb. Co-authored-by: Isaac <no-reply@databricks.com>
|
Thanks for the review! Fixes are in a057fdd, plus two stacked follow-ups: #7692 (native-line throttle) and #7691 (shared CI-env helper for the acceptance scripts). Nits from the review body: Q1 (split): the Q2 (single-agent table): fair point; with one agent it's more structure than strictly needed. I kept it to avoid more churn on this PR, and I'm happy to collapse it to a direct Claude Code check in a follow-up if you'd prefer. Q3 (CI and terminal checks): we measured Claude Code's Bash tool: stdin is |
…d install #7585 made the hint ask DatabricksTooling, which already walks Claude's project skills dirs, so the false "skills are not installed" warning an agent hit right after `aitools install --scope project` is fixed. The hint tests cover project-scoped plugins and HOME skills but not this layout, so add regression tests for a project install (relative symlink from .claude/skills into .databricks/aitools/skills) seen from the project root and from a subdirectory of the repository. Co-authored-by: Isaac <no-reply@databricks.com>
…d install #7585 made the hint ask DatabricksTooling, which already walks Claude's project skills dirs, so the false "skills are not installed" warning an agent hit right after `aitools install --scope project` is fixed. The hint tests cover project-scoped plugins and HOME skills but not this layout, so add regression tests for a project install (relative symlink from .claude/skills into .databricks/aitools/skills) seen from the project root and from a subdirectory of the repository. Co-authored-by: Isaac <no-reply@databricks.com>
## Release v1.27.0 ### Notable Changes * `databricks aitools install`, `update` and `uninstall` now exit with code 1 when they do nothing they were asked to. `install` fails when it sets up nothing: no coding agent is detected, every agent is skipped or fails to install, or no skill in the release can be installed (also with `--path`). `update` fails when every update or check it attempts fails, and `uninstall` when every removal it attempts fails. Scripts that ran these commands as a best-effort step and relied on exit code 0 should handle the new exit code. ([#7481](#7481)) * Add `file:<path>` selectors to `pipelines run` and pipeline `bundle run`. `--refresh` and `--full-refresh` now use space-separated lists, such as `--refresh "orders payments"`, instead of comma-separated lists. Single names containing spaces now require backticks, e.g. ``--refresh 'table:`order details`'``. ([#7184](#7184)) * The Databricks CLI is now available on PyPI as `databricks`: run it with `uvx databricks`, or install it with `uv tool install databricks`, `pipx install databricks` or `pip install databricks`. The package contains only the prebuilt CLI binary. Uninstall the legacy `databricks-cli` package first, since both install a `databricks` command. ([#7704](#7704), [#7777](#7777)) ### CLI * Add `databricks sandbox fs` commands (`cp`, `ls`, `cat`, `mkdir`, `rm`) to transfer files between your local filesystem and sandboxes using `<sandbox-id>:/path`. ([#7372](#7372)) * `databricks fs` commands retry reads (stat, list, open) and deletes on DBFS and Volumes when the server returns a transient error or does not respond within the HTTP timeout. ([#7556](#7556)) * Retry UC Volumes file operations (`fs cp`, `fs ls`, `fs rm`, bundle uploads to Volumes) when the Files API returns HTTP 408, 429, 502, 503 or 504 (for example "AWS S3 is throttling requests"), or on a transient network error; this regressed in v1.10.0. ([#7559](#7559)) * Stop warning when the host metadata request (`/.well-known/databricks-config`) times out; the CLI already falls back to your config, and the message is still shown with `--debug`. ([#7752](#7752)) * Add `databricks aitools auto-install disable` and `enable` to turn off (or back on) the automatic installation of Databricks AI tools on CLI upgrades. A global-scope `databricks aitools uninstall` also turns it off. ([#7701](#7701)) * Stop printing the Databricks AI tooling install hint on `auth`, `configure`, `current-user me`, `version`, `help`, and shell-completion commands. ([#7452](#7452)) * When Codex, Cursor, GitHub Copilot CLI, OpenCode, Gemini CLI, Antigravity, Pi, Goose or Kiro run the CLI without Databricks skills installed for that agent, the CLI now prints a one-line hint to install them with `databricks aitools install --agents <agent> --scope global`, previously shown only to Claude Code. ([#7586](#7586)) * `databricks aitools install` now rejects an unknown `--agents` name before it shows the scope prompt, instead of after. ([#7469](#7469)) * `databricks aitools install` no longer asks "Proceed?" after the agent picker; the picker selection confirms the plan. The picker title now shows the install scope, its hint says Enter installs, `--skills-only` runs label plugin agents "skills" and pre-check every agent whose config dir exists, and the plan summary after it is reduced to the scope line. ([#7478](#7478)) * `databricks aitools install --output json` reports more specific `error_category` values: `SKILLS_FETCH_FAILED`, `RELEASE_LOOKUP_FAILED`, `RATE_LIMITED`, `MANIFEST_INVALID`, `STATE_INVALID` and `PLUGIN_CLI_TIMEOUT`. A coding agent CLI that does not answer in time is now reported as `timed out after <duration>`, followed by any stderr it wrote before it was stopped. ([#7475](#7475)) * `databricks aitools install` no longer prompts when a coding agent runs it, even in a pseudo-terminal; it uses the non-interactive defaults instead and prints a note saying so. ([#7477](#7477)) * `databricks aitools install` now prints the directory it installed skills into (for example `Installed 1 skill in ~/.databricks/aitools/skills.`), so agents can open the installed `SKILL.md` files without searching for them. ([#7668](#7668)) * `databricks aitools install` now ends a failed install with a step to recover, in the text output and in a new `recovery` field of `--output json`. ([#7480](#7480)) * `databricks aitools install` now installs plugins for the other coding agents even when installing skill files fails. ([#7479](#7479)) * `databricks aitools install` now installs skill files for a detected coding agent whose CLI is not on PATH (for example a Claude Code desktop or IDE install), unless it already has the plugin, says how to switch to the plugin later, and warns when the current project also has the plugin. `databricks aitools update` no longer tries to install the plugin for such an agent, and no longer writes skill files next to a plugin that such an agent already has. ([#7482](#7482)) * Fix `databricks aitools install` waiting a long time on an unreachable or slow workspace while it resolves authentication for telemetry; that wait is now limited to a few seconds. ([#7468](#7468)) * `databricks aitools install`, `update` and `list` now refuse a skills manifest that names a skill or file outside the skills directory, before writing any skill files. A skill whose name or file paths are not portable is skipped with a warning, and `update` keeps its installed copy. ([#7722](#7722)) * `databricks aitools install` and `databricks aitools update` refresh only the marketplace the Databricks plugin is listed in, so another unreachable marketplace no longer blocks them. When that refresh fails, both commands warn, install or update the plugin from the local copy of the marketplace, and record the plugin version as unknown. ([#7473](#7473)) * `databricks aitools install` and `databricks aitools update` now wait up to 20 seconds instead of 5 for a coding agent's CLI to answer the plugin capability check, so a slow first start is less likely to report the agent as unavailable. ([#7472](#7472)) * `databricks aitools install` no longer fails a plugin-only install when the lookup of the latest skills release on GitHub fails or is rate-limited. The plugin is installed and its version is recorded as unknown. ([#7474](#7474)) * `databricks aitools` commands now send usage telemetry, and the CLI records when it recommends installing Databricks AI tools. When the configured workspace is slow or unreachable, `aitools install`, `update` and `uninstall` can take up to about 6 seconds longer, and `list`, `version` and runs that stop early up to about 3 seconds. Set `DATABRICKS_CLI_DISABLE_TELEMETRY=1` to opt out. ([#7467](#7467), [#7469](#7469), [#7470](#7470)) * `databricks aitools install` and `databricks aitools update` now report a coding agent CLI shim (`.cmd` or `.bat`) on Windows that prints its plugin help and then exits with an error as `shim-exited-with-error`, with the command to run to see the error, instead of reporting the CLI as not on PATH. ([#7476](#7476)) * `apps init` no longer suggests installing skills when every detected coding agent already has the Databricks plugin. ([#7467](#7467)) * `databricks auth login` now gives up on the optional workspace lookup after about 5 seconds instead of retrying for about 5 minutes, and explains the fallback to the manual workspace ID prompt. ([#7522](#7522)) * Throttle Claude Code's native plugin install hint separately from the text hint: it prints at most once per 10 minutes per session, so the install dialog still comes up when a subagent, a hook, or a command outside a Claude Code subprocess runs the session's first `databricks` command. ([#7692](#7692)) * Print Claude Code's native plugin install hint alongside the text hint, so Claude Code offers a one-click install of the Databricks plugin. Claude Code now counts as having the tooling only when it loads it: skills only in `~/.databricks/aitools/skills` don't count, and a project- or local-scope plugin counts only when the CLI runs inside that project. `auth token`, `auth docker token` and CLI subprocesses no longer print install hints. ([#7585](#7585)) * Interactive prompts and pagers now work in Git for Windows' bash and other MSYS2 shells when they run under ConPTY (the default since Git for Windows 2.47) or directly in a Windows console. ([#7699](#7699)) * Add the `databricks feedback` command to send feedback about the CLI to Databricks. ([#7724](#7724)) * Fix shell completion in `databricks fs` commands for local paths typed with a leading `./` or repeated separators, which previously offered no suggestions. ([#7679](#7679)) * Fix a panic in `databricks labs` commands when a non-string SDK setting such as `DATABRICKS_DEBUG_HEADERS` or `DATABRICKS_RATE_LIMIT` is set in the environment. ([#7673](#7673)) * Fix a panic in `databricks labs install` for projects whose `labs.yml` has no `install` section. ([#7673](#7673)) * `NO_COLOR` no longer disables prompts, spinners or the list pager; they run without color. ([#7700](#7700)) * Add the Beta `databricks pipelines test` command, which runs the tests of a pipeline, and the `testing` property of pipelines in bundles. ([#7429](#7429)) * Add `databricks sandbox forward`, which serves an app running in a sandbox at a local URL. ([#7395](#7395)) * Bound the whole telemetry upload at exit, including config resolution and authentication, by its timeout, so a stalled auth or telemetry endpoint no longer keeps a command from exiting. ([#7467](#7467)) * `databricks version --check` now suggests `databricks aitools install` on stderr when a coding agent is detected without the Databricks plugin or skills. ([#7471](#7471)) ### AI Runtime * `air get -o json` now reports `PENDING` while the run waits for compute, matching `air list`, and includes the run's `status_message`. ([#7455](#7455)) * Wrap long failure reasons and log lines in the `air list` info and logs panes, keep the MLflow link (labeled with its retry) when a run's last retry failed before starting, and point `air logs` at earlier retries when the last one has no logs. ([#7456](#7456)) * `databricks air logs --node` now reports an out-of-range node index when it can determine the run's node count, instead of looking like a successful run with no logs. ([#7516](#7516)) * Add experiment, duration, accelerator, and dashboard metadata to `air list --output json`. For retried runs, `started_at` now reports the latest attempt, matching `air get`. ([#7568](#7568)) * `air list` shortens MLflow run names and unnamed-run IDs by keeping their end behind a leading `…`, and its columns no longer overflow on wide (for example CJK) text; `air get` shows an unnamed run's full MLflow run ID. ([#7582](#7582)) * Shrink the Experiment and User columns of the interactive `air list` table to fit the terminal width, keeping the end of a shortened experiment name. ([#7583](#7583)) * Show the experiment, accelerators and configuration in the `air list` info pane for runs where it showed `N/A`. ([#7587](#7587)) * Strip terminal control characters from the MLflow run, experiment and user names shown by `air list` and `air get`, so a run name cannot clear or retitle the terminal of whoever lists it. ([#7601](#7601)) * `air logs`, `air get`, and `air run --watch` no longer make an extra credential-check request before fetching the run, and `air get` reuses the configured workspace ID instead of looking it up. ([#7613](#7613)) * `air list` no longer waits on extra current-user requests before listing runs, which makes the default list about 150ms faster. ([#7615](#7615)) * `air list --all-status` requests larger index pages, so it makes far fewer round trips once the server supports `page_size`. ([#7616](#7616)) * Fit the interactive `air list` table to narrow terminals: when the Experiment and User columns are at their floors, shorten Started to `MM-DD HH:MM` and size the MLflow column to its labels, then hide Accelerators, User, MLflow and Progress in that order, naming the hidden columns in the hint line. ([#7629](#7629)) * `air get -o json` now includes `mlflow_run_id` and `mlflow_experiment_id`, so a run can be handed to MLflow-based steps without parsing `mlflow_url`. They identify the MLflow run `mlflow_url` links to, including one from an earlier retry that `mlflow_retry` names. ([#7636](#7636)) * `air get` and the `air list` info pane now show the environment of runs on the Databricks AI environment (`databricks_ai_v*`) instead of `N/A`. ([#7637](#7637)) * `air run` now rejects a `code_source.include_paths` path that selects no files instead of uploading a snapshot without it; glob patterns and pathspec magic are left to git. ([#7643](#7643)) * `air convert-to-dabs` now suggests the `--output-dir` that converts a working-tree workload whose `code_source.root_path` is outside the bundle directory, instead of suggesting `sync.paths`. ([#7644](#7644)) * `air run` now lists a `code_source.root_path` outside a git repository relative to `root_path`, so running from a subdirectory of `root_path` no longer uploads an empty or partial snapshot. ([#7646](#7646)) * `air list --all-status` now fills `--limit` past purged or inaccessible indexed runs. ([#7743](#7743)) * `air run` now warns about missing PyTorch when a pinned snapshot's `include_paths` selects a Python file. ([#7744](#7744)) * `air cancel JOB_RUN_ID` now rejects run IDs that are not AI Runtime runs instead of cancelling unrelated Jobs runs. In `-o json` output, the `failed[].error` text for a run that does not exist or that you cannot view now starts with `failed to get run <id>:` instead of the cancel API error. ([#7569](#7569)) * `air get -o json` now reports `mlflow_retry` when the MLflow link comes from an earlier retry because the run's last retry never reached MLflow. ([#7620](#7620)) * `databricks air run --help` now shows how to install the databricks-ai-runtime coding agent skill. ([#7652](#7652)) * `air run --watch -o json` now reports each lifecycle `STATUS` transition once and never reports a non-terminal state after the run has ended, including when logs are served from MLflow artifacts. ([#7621](#7621)) * `air convert-to-dabs` now explains the `--output-dir` workaround when `code_source.root_path` resolves to the bundle root, and notes when the generated job carries the default `max_retries`. ([#7579](#7579)) * `air run` now refuses to upload a `code_source` snapshot that selects no files (for example when `root_path` is git-ignored) and explains the cause, and `--dry-run` runs the snapshot file selection and reports the file count. ([#7570](#7570)) * Repeat the Jobs run link after live log streaming ends. ([#7571](#7571)) * `air get -o json` includes run details including flat `accelerator_type` and `num_accelerators`. Use `--include-config` to include stored config, which may contain plaintext `env_variables` and parameters; downloads over 1 MiB are omitted, and config-dependent `max_retries` is otherwise null. ([#7624](#7624)) * `air get` and the `air list` info pane no longer show `Max Retries 0` when the Jobs API leaves the task's retry limit out. For a run submitted with `air run`, the limit now comes from the run's uploaded config, shown as `3 (CLI default)` when the config doesn't set `max_retries`; it shows `N/A` when the limit can't be known (job runs, multi-container runs, legacy runs with no explicit value, or an unreadable config). ([#7623](#7623)) * Reorganize `databricks air` help: subcommands are grouped by task, `air run` has examples, and YAML fields are listed under headings, with pool fields marked optional and `command`, `compute.num_accelerators`, and `compute.accelerator_type` correctly marked required. ([#7512](#7512)) * `air run -h config` now lists the environment variables AI Runtime sets on every node (such as `MLFLOW_RUN_ID` and `CODE_SOURCE_PATH`), and `air logs --help` notes that `--download-to` cannot be combined with `--tail` or `--minutes`. ([#7655](#7655)) * Support full workspace accelerator types in `air list` filters while preserving family matching, and reject invalid filter values. ([#7515](#7515)) * `databricks air list --all-status` now warns when the AI Runtime run index is unavailable and it falls back to scanning Jobs runs. That fallback scan is capped at 250 runs, and a stuck index request fails after 15 seconds instead of retrying for minutes. `--all-status --filter user=<your email>` now uses the run index instead of scanning the workspace's full run history. ([#7619](#7619)) * Return structured JSON errors for invalid `air list --filter` values when `-o json` is used. ([#7612](#7612)) * `air list --all-status` no longer fails when a listed run has been purged. ([#7523](#7523)) * **Behavior change:** `databricks air logs` now prints the logs available now and exits, also for an active run, matching `kubectl logs`, `docker logs` and `databricks apps logs`. Pass `--follow` (`-f`) to stream new lines until the run completes; `--tail N --follow` prints the last N lines first. `databricks air run --watch` still submits and streams. ([#7658](#7658)) * `databricks air logs -o json` and `databricks air run --watch -o json` now emit a `WAITING` event, for example `{"type":"WAITING","message":"Waiting for accelerator compute capacity..."}`, when the reported waiting status changes. ([#7695](#7695)) * `databricks air logs` on a run that ended in the last two minutes now waits for the log index to settle before printing, instead of sometimes dropping lines that were indexed late. ([#7663](#7663)) * `air logs` and `air run --watch` now print the waiting status (for example `Waiting for accelerator compute capacity...`) as a line on stderr when stderr is not a terminal, instead of printing nothing until the run produces logs. ([#7657](#7657)) * Route AIR log links to the MLflow Logs page instead of the hidden artifact path. ([#7491](#7491)) * Invalid `compute.accelerator_type` errors and `air run -h config.compute` no longer list the CLI's built-in accelerator types, which could differ from the workspace's. ([#7520](#7520)) * `databricks air --help` now notes that GPU pools are not required to run workloads. ([#7581](#7581)) * `air run` now prints the job run URL with the `?o=<workspace-id>` suffix, matching `air get` and `air list`; `dashboard_url` in `air run -o json` and the `--watch` events gains it too. When `air list` cannot resolve the workspace ID, its links now omit `?o=` instead of ending in `?o=0`. ([#7573](#7573)) * `databricks air run` now reports an unknown workload YAML key by its field path, suggests the closest valid field, and points to `-h config` instead of naming an internal Go type. ([#7574](#7574)) * `air run` (including `--dry-run`) now warns before submitting when the `code_source` imports `torch` but PyTorch is not declared: the environment is the Standard environment (version 5 or later) with no `torch` dependency, so the run may fail after it waits for a GPU. ([#7656](#7656)) ### Bundles * Add `isolation_mode` to the `catalogs` resource, so a catalog can be deployed as `OPEN` or `ISOLATED`. ([#7463](#7463)) * `bundle validate` no longer warns when resource permissions and top-level or target permissions name the same principal. ([#7291](#7291)) * Retry bundle file uploads, reads and deletes when the server does not respond within the HTTP timeout ("request timed out after ... of inactivity"). ([#7521](#7521)) * Show how to recover when `bundle deploy` fails because an app's service principal does not exist or was deleted. ([#7533](#7533)) * Retry transient server errors when creating Lakebase (`postgres_*`) resources in the direct deployment engine. ([#7557](#7557)) * Fix `bundle plan` reporting the `spark_conf` and `custom_tags` the backend derives from `is_single_node: true` as drift on clusters. ([#7577](#7577)) * Stop planning a spurious update for Lakeflow Connect pipelines whose `ingestion_definition.data_staging_options` is filled in by the backend. ([#7603](#7603)) * Retry the root path check in `bundle destroy` on transient errors, so a dropped connection no longer fails the destroy. ([#7597](#7597)) * Use the Lakebase branch when its create request times out but the branch was created, instead of failing the deploy, in the direct deployment engine. ([#7602](#7602)) * Retry transient server errors when deleting resources in the direct deployment engine. ([#7708](#7708)) * Use the Lakebase project when its create request times out but the project was created, instead of failing the deploy with "already exists", in the direct deployment engine. ([#7747](#7747)) * Use the feature, model serving endpoint or dashboard that a create request created despite timing out, instead of failing the deploy, in the direct deployment engine. ([#7757](#7757)) * Use the Lakebase role when its create request fails with a transient error but the role was created, instead of failing the deploy, in the direct deployment engine. ([#7767](#7767)) * Use the MCP service, model service or model provider service that a create request created despite failing with a timeout, a 499 or a reset stream, instead of failing the deploy, in the direct deployment engine. Create recovery for features, serving endpoints and dashboards now also covers 499 and reset streams. ([#7782](#7782)) * Retry app updates and deletes in the direct deployment engine when they fail with 400 BAD_REQUEST because the app compute is `UPDATING`, after waiting for the compute to settle. ([#7606](#7606)) * Enable drift detection for `source_code_path` and `git_source` on apps. ([#6399](#6399)) * `bundle destroy` and `bundle deploy` no longer fail with "internal error, missing in state" when the state holds an empty-ID entry left by an interrupted recreate in an older CLI. ([#7676](#7676)) * Fix a feature whose `full_name` contains uppercase letters being recreated on every deploy. ([#7398](#7398)) * Fix `bundle deployment bind` (and `bundle generate --bind`) failing with `no such node "resources.internal_immutable_snapshots.immutable"` when `workspace.sync_mode` is `immutable`, as reported in [#7306](#7306). ([#7329](#7329)) * Retry creating the deployment lock (without `--force-lock`) when the request fails with a transient server error, times out ("request timed out after ... of inactivity") or loses its connection ("unexpected EOF"). ([#7594](#7594)) * Fix a materialization set member whose `feature_name` contains uppercase letters failing to deploy, or recreating the set on every deploy. ([#7398](#7398)) * Delete the old resource when its name follows a renamed schema. ([#7718](#7718)) * Fix `pipelines dry-run` starting a real run when given a job KEY, and a crash in its shell completion. ([#7672](#7672)) * Rename schemas in place; schemas and volumes follow a renamed catalog or schema. ([#7343](#7343)) * Fix a panic when validating or deploying jobs with local libraries and serverless job clusters. ([#7725](#7725)) * Fix `tgz` artifacts with `git` set producing an empty tarball or failing when the bundle is in a subdirectory of the git repository. ([#7674](#7674)) * Apply `presets.trigger_pause_status` and the development mode pause default to `triggers[]` entries on jobs. ([#7450](#7450)) ### Dependency Updates * Bump dependencies with known vulnerabilities. ([#7750](#7750)) * Update the Go SDK to [v0.198.0](https://github.com/databricks/databricks-sdk-go/releases/tag/v0.198.0). ([#7530](#7530)) * Update the Go SDK to [v0.199.0](https://github.com/databricks/databricks-sdk-go/releases/tag/v0.199.0). ([#7584](#7584)) * Update the Go SDK to [v0.200.0](https://github.com/databricks/databricks-sdk-go/releases/tag/v0.200.0). ([#7653](#7653)) * Update the Go SDK to [v0.201.0](https://github.com/databricks/databricks-sdk-go/releases/tag/v0.201.0). ([#7769](#7769))
databricks#7692) ## Changes The native `<claude-code-hint>` line now has its own throttle (a separate cache bucket, 10-minute TTL), keyed by the hint's throttle key from databricks#7586 and written only when the line actually prints. The text hint keeps its one-hour marker, and other agents' durable claim (`claimHintMarker`) now goes through the same text throttle. The native throttle applies only to Claude Code, the one agent with a native line. All throttles are still checked before the plugin/skills filesystem probe. The `CLAUDE_CODE_NATIVE` hint-shown event stays on the text hint's hourly clock. It is logged only for the first native line printed since the text line last printed, so a native-only repeat logs no event. ## Why This addresses simonfaltum's review comment on `hint.go` in databricks#7585. The single shared marker was written whenever the text hint printed. That included commands where the native line was withheld (no `CLAUDE_CODE_CHILD_SESSION`, CI, a terminal attached), and commands where Claude Code strips the tag without prompting (subagents, hooks). After that, the main loop got neither line for an hour. The CLI can't tell a subagent from the main loop, because both carry the same session id and child-session marker. So a separate key alone doesn't fix the subagent case; the shorter TTL is what bounds it. Claude Code shows the dialog at most once per session, so repeating the line is cheap: a raw tag reaches a `2>&1 | jq` reader at most once per 10 minutes. Telemetry costs nothing extra, because a native-only repeat logs no event and so triggers no exit-time upload (`hintOnlyUploadTimeout`). Native-event counts stay at most one per text window instead of growing with session length. In return, a session waits at most 10 minutes for the dialog. ## Tests - New unit tests: - A withheld native line (no child session, CI) leaves the next command's native line due. Only that line prints, and it logs its event. - The native line reappears after its TTL while the text hint stays throttled, and that repeat logs no event. A new text window logs both events again. - A text-only print starts a new event window, so the next native print logs its event. - New `acceptance/cmd/aitools/claude-config-dir` case for the withheld-then-native sequence. The existing throttle case's title now describes both clocks. - `libs/telemetry/protos/README.md` describes both clocks and the hourly native event. - `go test ./libs/aitools/... ./cmd/root/...`, `go test ./acceptance -run 'TestAccept/cmd/aitools'`, `./task test-exp-aitools`, `./task checks` and `./task lint-q` all pass. This pull request and its description were written by Isaac. --------- Co-authored-by: Isaac <no-reply@databricks.com>
## Changes Adds an `unset_ci_env` helper to `acceptance/script.prepare`. It replaces the three copies of the CI and CI/CD provider `unset` line in `cmd/aitools/claude-config-dir`, `cmd/aitools/telemetry/hint` and `air/image-push`. The helper's comment points at `listCiCdProviders` in the Go SDK's `useragent/cicd.go` so the list stays in sync. The helper is also listed in the built-in helpers section of `.agents/rules/testing.md`. ## Why Addresses simonfaltum's review comment on databricks#7585 about the duplicated line. The helper lives in the root `script.prepare` because that is the only prepare file that covers both `cmd/aitools` and `air`. After this change, `unset_ci_env` is the only copy of the list. ## Tests - Rebased on current main, which includes databricks#7692. `go test ./acceptance -run TestAccept/cmd/aitools` (14 dirs) and `-run TestAccept/air` (24 dirs) both pass with no golden changes. They pass in a clean env and with `CI=true GITHUB_ACTIONS=true` exported. - Negative control: I turned the helper into a no-op and exported the CI vars. `cmd/aitools/telemetry/hint`, `cmd/aitools/claude-config-dir` and `air/image-push` all fail. This pull request and its description were written by Isaac. --------- Co-authored-by: Isaac <no-reply@databricks.com>
…d install (databricks#7667) ## Changes Test-only: adds `TestMaybeHint_ProjectSkillsInstalledSilent` and `TestMaybeHint_ProjectSkillsInParentOfRepoSubdirSilent`. They lay out an install the way `databricks aitools install --scope project` does (a relative symlink from `.claude/skills/<skill>` to the project's `.databricks/aitools/skills/<skill>`). Then they check that the "Databricks skills are not installed" hint stays quiet, both from the project root and from a subdirectory of the repository. ## Why A coding agent in an AI Runtime trial ran `databricks aitools install --scope project`. The very next command, `databricks air list`, said skills were not installed and suggested reinstalling. This PR originally fixed that in `claudeToolingInstalled`. databricks#7585 has since made the hint ask `DatabricksTooling`, which already walks Claude's project skills folders, so the fix is no longer needed. The hint tests cover project-scoped plugins and skills under HOME, but not a project-scoped skills install, so I kept the regression tests. Both fail on the commit before databricks#7585. ## Tests `go test ./libs/aitools/...` passes. This pull request and its description were written by Isaac. --------- Co-authored-by: Isaac <no-reply@databricks.com>
Changes
When Claude Code runs the CLI without the Databricks plugin or skills, the install hint now also prints Claude Code's native
<claude-code-hint>line, which makes Claude Code offer a one-click install ofdatabricks@claude-plugins-official.Agent.DatabricksTooling: skills only in the shared~/.databricks/aitools/skillsstore (or the legacy~/.databricks/agent-skills) no longer count, and a plugin recorded withprojectorlocalscope counts only when the CLI's working directory is inside the recorded project.CIcounts as unset for the same false valuesenv.GetBoolaccepts (0,false,no,off, ...); any other value, such as a provider name, means CI.cmdio.HasTerminal(any standard stream is a TTY) andcmdio.FakeTTYReaderfor tests.agents.SelfExecEnvis set on the child processes ofair images push(auth docker configure) andapps init(post-create deploy/dev runs).protos/README.md,aitools_events.go): oneaitools_hint_eventper printed hint line..nextchanges/cli/claude-code-native-hint.md.Why
The text hint asks the model to run
databricks aitools install; the native line lets Claude Code offer the user the plugin install directly. Claude Code shows that dialog only in an attended session, at most once per session and once ever per plugin, and strips the tag before the model sees the output. The native line is gated so a human never sees the raw tag: it needsCLAUDE_CODE_CHILD_SESSION(IDE terminals don't set it), no terminal on any standard stream (tmux or a shell started from inside Claude Code inherit the env), and no CI.auth token,auth docker tokenand CLI-spawned CLI processes are skipped because their stderr is embedded mid-line in another tool's error, where Claude Code can't strip the tag.Detection follows what Claude Code itself loads, so the hint fires exactly when Claude Code has no Databricks tooling. Claude Code loads project and local plugins from the directory it started in, which the CLI can't see, so the CLI's cwd stands in for it. As a result a session started in
~/repowith a project-scoped plugin that runscd /tmp/x && databricks ...now gets the hint; before this PR any plugin record kept it quiet.Door: two-way. Pure CLI output and telemetry change; reverting removes the line. No exit code changes and no extra network calls.
Blast radius: small.
2>&1 | jqpipeline in that command reads it raw. With the file cache disabled, both lines print on every eligible command.~/.databricks/aitools/skillsor~/.databricks/agent-skillsno longer silence the hint for Claude Code, and project/local plugins silence it only inside their project.auth token,auth docker tokenand CLI-spawned CLI processes no longer print any install hint.CLAUDE_CODE_NATIVE, is logged.Tests
Run with the host's agent and CI env scrubbed.
MaybeHintremoved, tests unchanged):TestAccept/air/image-push: inside a simulated Claude Code session, theauth docker configurechild adds no hint to the parent's output or to its error text.TestMaybeHint_NativeHintalso covers: CI (CI=true,CI=<provider name>,TF_BUILD) prints the text only,CI=false/no/OFFstill print the native line; plugin or Claude skills installed prints nothing;auth token,auth docker token, self-exec and aitools commands print nothing.TestMaybeHint_ProjectPluginCountsOnlyInItsProject: a project-scoped plugin silences the hint only inside its project.TestMaybeHint_NativeHintSilentOnTerminal: a terminal on stdin, stdout or stderr drops the native line.This pull request and its description were written by Isaac.