Skip to content
Draft
Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
14 commits
Select commit Hold shift + click to select a range
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
11 changes: 10 additions & 1 deletion PRIVACY.md
Original file line number Diff line number Diff line change
@@ -1,7 +1,7 @@
# Privacy Policy — Altimate Code plugin for Claude Code

**Effective date:** November 2, 2026
**Last revised:** November 19, 2026
**Last revised:** November 20, 2026
**Applies to:** the `altimate-code` Claude Code plugin distributed from this repository, and the data flows it initiates through the [altimate-code](https://github.com/AltimateAI/altimate-code) MCP server and CLI.

This policy supplements the Altimate AI [Privacy Policy](https://www.altimate.ai/privacy) and [Terms](https://www.altimate.ai/terms), which govern Altimate services generally. Where this document is silent, the master policy applies. Questions: **info@altimate.ai**.
Expand All @@ -17,6 +17,7 @@ This policy supplements the Altimate AI [Privacy Policy](https://www.altimate.ai
- Warehouse queries, dbt commands, and file edits run **locally on your machine, under your credentials**. Your warehouse credentials are never sent to Altimate.
- In a project linked to an Altimate workspace, the plugin copies the workspace's team skills into the project's `.claude/skills/` and refreshes them from Altimate in the background at session start and at most hourly. Their text loads into the Claude Code conversation like any project skill. `ALTIMATE_CLAUDE_SKILL_SYNC=0` turns this off and removes the copies at the next session start.
- The plugin counts its own use anonymously, through the altimate-code CLI's usage telemetry: one event at each session start (plugin and altimate-code versions, OS, a random install id) and one each time a panel or command opens or a gate holds or allows a command (its name and outcome). No session id, project, model, column or file names, SQL or session text, and nothing that links them to your account or to the CLI's other telemetry. `ALTIMATE_TELEMETRY_DISABLED=true` turns it off. See [Plugin usage events](#plugin-usage-events).
- `/altimate-feedback` only drafts a GitHub issue and shows it to you; nothing is sent unless you then confirm with `--confirm` (see [Feedback](#feedback)).
- Installing the plugin is an explicit act — `/plugin marketplace add` does not install it, and `/plugin install` never auto-runs. Nothing is delegated until the plugin is installed and you send a task that matches the skill.

## What the plugin does and does not do
Expand Down Expand Up @@ -108,6 +109,14 @@ To learn whether the report gets used, the plugin asks the local altimate-code C

Each event carries the step and the Claude Code session id (`claude:<id>`) only: no model, column, file or project names. It goes through Altimate's usage telemetry, described in [Usage telemetry](#usage-telemetry), and that section's opt-out turns it off; with `ALTIMATE_TELEMETRY_DISABLED` set, the plugin does not start the CLI for it. With an altimate-code that predates the command, nothing is sent.

## Feedback

`/altimate-feedback` drafts a GitHub issue for this plugin's repository. **Nothing is sent when you run it**: it shows you the exact title and body, and only `/altimate-feedback --confirm` sends them. The draft is kept in memory until then and is dropped when you discard or replace it, or the session ends. The plugin does not start `gh` or open a browser before you confirm.

The issue holds your own words (with the dbt profile's secret values, tokens, passwords, connection-string credentials and paths under your home folder cut out) and these facts: the plugin's version, altimate-code's version and binary name (not its path), the Claude Code version, the OS, dbt-core, adapter and Python versions when the dbt doctor has run, and the plugin's on/off settings. With `--with-error` it also holds the kind of the last error the plugin met: only the error's type and code (for example `Database Error`, `ENOENT`) and a few fixed phrases are kept, none of the message, so no file path, SQL, or project, model, relation or column name comes with it. The plugin collects no file paths, SQL, or project, model or column names; if you write them in your description they are included (a home-folder path is cut out when it is quoted or has no space in a folder name), so check the draft.

After you confirm, the text goes to GitHub: through `gh issue create`, signed in as you, or through a prefilled link you open and submit yourself. The issue is public on [AltimateAI/altimate-claude-plugin](https://github.com/AltimateAI/altimate-claude-plugin/issues) and subject to GitHub's terms and privacy statement. Altimate receives it as any public issue; the plugin sends nothing to Altimate's services for this command. Before creating the issue with `gh`, the plugin runs `gh auth status` and `gh label list` for that repository, which contact GitHub with your sign-in but carry none of the draft. The plugin never signs you in to `gh`.

## Plugin usage events

To count installs, active use and which features are used, the plugin asks the local altimate-code CLI (`altimate-code tool report-event`) to send these events through Altimate's usage telemetry, described in [Usage telemetry](#usage-telemetry). The plugin itself contacts no server for them.
Expand Down
3 changes: 3 additions & 0 deletions plugins/altimate-code/CHANGELOG.md
Original file line number Diff line number Diff line change
Expand Up @@ -13,6 +13,9 @@ All notable changes to this plugin will be documented in this file.
- `--defer --state …`, `--favor-state` and `--target …` that `/defer` and `/target` add to Claude's dbt commands are no longer invisible: Claude's result says what the command ran as, and a line above the prompt shows it until the next prompt.
- **Resolve** on a thread the lagging comment list still shows after it was resolved elsewhere is done (the thread is hidden) instead of answering "not an open thread"; a reply to such a thread says it is no longer open and reads the list again.

### Added
- `/altimate-feedback <what happened>` drafts a GitHub issue for this plugin's repository from your words and a few facts the plugin knows (plugin, altimate-code and Claude Code versions, the altimate-code binary name, OS, dbt-core, adapter and Python versions when the dbt doctor has run, and the plugin's on/off settings), shows the exact title and body, and creates it only after `--confirm`: with `gh issue create` (labelled `feedback` when the repository has that label) when `gh` is installed and signed in, otherwise as a prefilled new-issue link. `--with-error` also attaches the kind of the last plugin error (its type and code only, none of its message); `--cancel` discards the draft.

## [1.8.3] — 2026-11-19

### Fixed
Expand Down
13 changes: 13 additions & 0 deletions plugins/altimate-code/README.md
Original file line number Diff line number Diff line change
Expand Up @@ -156,6 +156,18 @@ Compile and validate runs a query on your warehouse after every model edit (`dbt

When Claude mentions `altimate-dbt`, it is the dbt command line that comes with altimate-code (`npm i -g @altimateai/altimate-code`, the same install as the plugin's engine). It finds your dbt project folder, Python environment and warehouse adapter by itself, and altimate-code's agent uses it instead of the raw dbt tool for better error messages. Its commands include `init`, `doctor`, `compile`, `build`, `run`, `test`, `deps`, `columns <model>`, `graph` (lineage) and `execute "<sql>"` (a query through the dbt adapter). You do not need to run it yourself: Claude and the altimate-code agent call it when a task needs it, and the plugin's own dbt helpers above do not depend on it. See [altimate-dbt in the altimate-code docs](https://docs.altimate.sh/data-engineering/tools/dbt-tools/#altimate-dbt-cli).

## Send feedback

**`/altimate-feedback <what happened>`** drafts a GitHub issue for this plugin's repository ([AltimateAI/altimate-claude-plugin](https://github.com/AltimateAI/altimate-claude-plugin/issues)) from your words and a few facts the plugin already knows, shows you the exact title and body, and creates the issue only after you confirm. Run it with no text and it tells you how to start.

1. `/altimate-feedback the band said nothing after a failed dbt run` shows the draft. Nothing is sent, and `gh` and the browser are not touched, yet.
2. `/altimate-feedback --confirm` creates exactly what you saw. With the [GitHub CLI](https://cli.github.com) installed and signed in (`gh auth status`), it runs `gh issue create` and prints the issue's URL, adding the `feedback` label when the repository has one. Without `gh`, or with `gh` not signed in or failing, it opens a prefilled new-issue link (and prints it, URL-encoded and kept under GitHub's length limit; a longer body is cut at the end and says so). Nothing is created until you press "Submit new issue" there. The plugin never signs you in.
3. `/altimate-feedback --cancel` discards the draft. Running the command again with new text replaces it. A draft is kept in memory only and ends with the session.

What the issue holds besides your words: the plugin's version, altimate-code's version and binary name (`altimate` or `altimate-code`, never its path), the Claude Code version, the OS, dbt-core, adapter and Python versions when the dbt doctor has run, and which of the plugin's on/off settings are on. Add `--with-error` to also attach the kind of the last error the plugin met: only its type and code (for example `Database Error`, `ENOENT`, `$.process.run`) and a few fixed phrases (`not found`, `timed out`) are kept, none of the message, so no path, SQL, relation, model or column name or secret can come with it. Your own words get the dbt profile's secret values, tokens, passwords, connection-string credentials and paths under your home folder cut out (a home path with a space in a folder name is cut whole when it is quoted or another separator follows the space); the rest is as you wrote it, so leave out what you do not want public: the repository's issues are public.

To see what would run without creating anything, set `ALTIMATE_FEEDBACK_DRY_RUN=1`: `--confirm` then prints the `gh issue create` command (or the link) and stops.

## Settings

The plugin has seven settings, shown when you enable it (`userConfig` in `.claude-plugin/plugin.json`). All are optional: Claude Code may report them as "not set", and each then takes the default below. `dbt_mods` is used by the dbt helpers, `maxScanBytes` by the scan check (BigQuery, Snowflake, Databricks), `healthBlocksPush` and `healthConfigPath` by the project health check, the two comments settings by team comments, and `show_edit_impact` by the per-edit impact note.
Expand Down Expand Up @@ -217,6 +229,7 @@ Still stuck? See [Support](#support).
- **Agent** — `agents/analyst.md` — read-only data analyst (MCP or CLI)
- **Command** — `commands/altimate.md` — `/altimate-code:altimate` slash command for explicit delegation
- **Command** — `commands/rules.md` — `/altimate-code:rules` to answer proposed team rules
- **Mod command** — `hooks/feedback.ts` — `/altimate-feedback` to draft and, once you confirm, file a GitHub issue
- **Hooks** — `hooks/hooks.json` — SessionStart capability description and workspace skill copy; Stop hook hourly background skill refresh; the team rules hooks (`hooks-handlers/workspace-rules.sh`): SessionStart, UserPromptSubmit and Stop, which call `altimate learn hook <event>` and never fail the session (the per-turn hooks run in the background; session start waits at most 15 seconds); opt-in PreToolUse `warehouse-guard.sh` (needs `ALTIMATE_ROUTE_GUARD=1`)

## Support
Expand Down
2 changes: 2 additions & 0 deletions plugins/altimate-code/hooks/dbt-mods-core.ts
Original file line number Diff line number Diff line change
Expand Up @@ -124,6 +124,8 @@ export function redactionForms(value: string): string[] {
const repr = part.replace(/\\/g, '\\\\').replace(/\n/g, '\\n').replace(/\r/g, '\\r').replace(/\t/g, '\\t')
forms.add(repr)
forms.add(repr.replace(/'/g, "\\'"))
// A single-quoted YAML scalar doubles its apostrophes.
forms.add(part.replace(/'/g, "''"))
}
return [...forms].filter(one => one.length > 0)
}
Expand Down
8 changes: 8 additions & 0 deletions plugins/altimate-code/hooks/dbt-mods.tsx
Original file line number Diff line number Diff line change
Expand Up @@ -7,6 +7,7 @@ import type { Args, EngineInterface, EventResult, On, PluginOptions, RenderEleme
import type { CheckView, DoctorCheck, DoctorView, FailureView, QueryResult, SavedQuery } from '../types'
import { later } from './altimate-cli'
import { openTracked } from './usage'
import { doctorFacts, noteProblem, type DoctorFacts } from './last-problem'
import { loadPaneSizesOnce, noteTerminal, paneTerminalColumns, paneRequest, sizeControls, type PaneSpec } from './pane-size'
import {
cellText,
Expand Down Expand Up @@ -761,6 +762,7 @@ async function recordRun($: EngineInterface, command: string, verb: string, sele
if (generation !== mineGeneration) return null
await update($, failureAtom, () => view)
const names = failures.map(one => one.name)
noteProblem(`dbt ${verb} failed`, failures[0]!.message)
$.ui.toast(`✖ dbt ${verb} failed · ${names.slice(0, 3).join(', ')}${names.length > 3 ? ` +${names.length - 3}` : ''} · Explain or Fix above the prompt`, { timeoutMs: 8000 })
return view
}
Expand Down Expand Up @@ -1779,6 +1781,12 @@ export async function dbtModsReadOrEditAfter($: EngineInterface, e: { tool: stri
* A conversation ended (a /clear): its notes, context, check and failed run are not the next one's.
* Called from the plugin's one session.end hook. A check still running lands nowhere.
*/
/** What the dbt doctor found for /feedback: the versions only, null before it has run. */
export async function dbtModsDoctorFacts($: EngineInterface): Promise<DoctorFacts> {
const view = await read($, doctorAtom)
return view ? doctorFacts(view.checks) : null
}

export async function endDbtMods($: EngineInterface) {
generation++
// A check of the conversation that ended answers nobody in the next.
Expand Down
10 changes: 10 additions & 0 deletions plugins/altimate-code/hooks/features/index.ts
Original file line number Diff line number Diff line change
Expand Up @@ -10,6 +10,8 @@
import type { Args, EngineInterface, EventResult, On, PluginOptions, RenderElement, ToolCallInput, ToolCallResult } from 'claude-code'

import { later } from '../altimate-cli'
import { feedbackSessionEnd, feedbackSessionStart, registerFeedback } from '../feedback'
import { noteProblem } from '../last-problem'
import { keepStatusParts, restoreStatusParts } from '../status-line'
import type { BashCall, EditedFile, FileCall } from '../tool-chain'
import {
Expand Down Expand Up @@ -168,6 +170,9 @@ export type FeatureContext = {
}

function failed($: EngineInterface, id: string, part: string, error: unknown) {
try {
noteProblem(`${id} ${part} failed`, String(error))
} catch {}
try {
$.ui.log(`${id} · ${part} failed: ${String(error)}`, { to: 'debug' })
} catch {}
Expand Down Expand Up @@ -218,6 +223,9 @@ export function registerFeatures(on: On, context: FeatureContext) {
try {
registerDbtMods(on, context)
} catch {}
try {
registerFeedback(on, context)
} catch {}
}

export async function featuresSessionStart($: EngineInterface, e: Args<'session.start'>) {
Expand All @@ -231,6 +239,7 @@ export async function featuresSessionStart($: EngineInterface, e: Args<'session.
await attempt($, 'docs', 'sessionStart', () => docsSessionStart($, e))
await attempt($, 'cteProfile', 'sessionStart', () => cteProfileSessionStart($, e))
await attempt($, 'dbtMods', 'sessionStart', () => sessionStartDbtMods($, e))
await attempt($, 'feedback', 'sessionStart', () => feedbackSessionStart($))
}

/**
Expand Down Expand Up @@ -305,6 +314,7 @@ export async function featuresSessionEnd($: EngineInterface, e: Args<'session.en
await attempt($, 'docs', 'sessionEnd', () => docsSessionEnd($, e))
await attempt($, 'cteProfile', 'sessionEnd', () => cteProfileSessionEnd($, e))
await attempt($, 'dbtMods', 'sessionEnd', () => endDbtMods($))
feedbackSessionEnd()
}

/** Each feature's band line, in order, with its id; a feature with nothing to say is left out. */
Expand Down
Loading
Loading