> For the complete documentation index, see [llms.txt](https://ai-os-and-trend-finder.gitbook.io/ai-os-and-trend-finder-docs/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://ai-os-and-trend-finder.gitbook.io/ai-os-and-trend-finder-docs/docs/commands.md).

# Command Reference

All scripts are run via `bun run <script>`. Prerequisites: Bun 1.x installed, `bun install` completed.

***

## Development

### `dev`

Start the local development server with hot reload.

```
bun run dev
```

* **Prerequisite**: `bun install`
* **Side effects**: Seeds `src/data/live-data.json` from the example file if missing. Opens a browser tab automatically.
* **Expected output**: Vite dev server running at `http://127.0.0.1:5189/`
* **Pass**: Server starts, page loads in browser
* **Fail**: Port conflict, missing dependencies, seed file copy failure

### `scripts/cleandev.sh`

Clean local generated state, reinstall dependencies, validate, hydrate local data, and optionally restart the tmux-managed dev server.

```
scripts/cleandev.sh
scripts/cleandev.sh --validate-only
scripts/cleandev.sh --no-start
```

* **Prerequisite**: Bun, tmux, curl, and `ss` or `lsof`
* **Side effects**: Stops project-owned dev processes, removes ignored generated state and `node_modules`, reinstalls from `bun.lock`, runs typecheck and tests, and refreshes host/local data unless `CLEANDEV_RUN_HOST_AGGREGATE=0` is set.
* **Pass**: Exit code 0, phase summary printed, aggregate log path printed when hydration writes one, and the default mode starts the app in `aios-dev`.
* **Validation-only**: `--validate-only` and `--no-start` run clean, install, typecheck, tests, and hydration, then mark start/runtime hints as skipped and leave the app down.
* **Extension wording**: Output reports UI extension state separately from extension collector state.
* **Fail**: Missing required commands, non-project port ownership, dependency install failure, typecheck/test failure, hydration failure, or runtime health failure in restart mode.

### `build`

Production build via Vite.

```
bun run build
```

* **Prerequisite**: `bun install`
* **Side effects**: Seeds live-data.json if missing. Writes output to `dist/`.
* **Expected output**: Vite build summary with chunk sizes
* **Pass**: Exit code 0, `dist/` directory created
* **Fail**: TypeScript or import errors in source

### `build:dev`

Development-mode build (unminified, with source maps).

```
bun run build:dev
```

* **Prerequisite**: `bun install`
* **Side effects**: Same as `build`
* **Expected output**: Vite build summary (development mode)
* **Pass**: Exit code 0
* **Fail**: Same as `build`

### `preview`

Serve the production build locally.

```
bun run preview
```

* **Prerequisite**: `bun run build` completed
* **Expected output**: Currently blocked by B010 because Vite preview expects `dist/server/server.js` while the Cloudflare Worker build emits `dist/server/index.js`
* **Pass**: After B010 is resolved, server binds to `127.0.0.1:4176`
* **Fail**: No `dist/` directory or the known B010 preview entry mismatch
* **Known gap**: Use `bun run worker:preview` on `http://127.0.0.1:8788` for full Worker SSR preview until B010 is resolved.

### `worker:preview`

Build and run the Cloudflare Worker build locally with the pinned Wrangler.

```
bun run worker:preview
```

* **Prerequisite**: `bun install`; runs `bun run build` first
* **Expected output**: Wrangler serves the Worker SSR build on `http://127.0.0.1:8788`
* **Pass**: Worker binds locally and serves the dashboard
* **Fail**: Build failure or Wrangler config/port error
* **Notes**: Uses the generated `dist/server/wrangler.json`. This is the current full-SSR preview path while the plain `preview` B010 gap is open.

### `worker:deploy`

Build and deploy the Cloudflare Worker build with the pinned Wrangler.

```
bun run worker:deploy
```

* **Prerequisite**: `bun install` and authenticated Wrangler credentials; runs `bun run build` first
* **Side effects**: Publishes the Worker to the configured Cloudflare account
* **Pass**: Wrangler reports a successful deploy
* **Fail**: Build failure, missing credentials, or Wrangler deploy error
* **Notes**: Deployment is outward-facing. See [Deployment](/ai-os-and-trend-finder-docs/docs/deployment.md) before running it.

***

## Quality

### Final certification pass

Run the final local hackathon certification in this order:

```bash
bun run typecheck
bun run typecheck:scripts
bun run test -- scripts/lib/apify/__tests__ scripts/lib/ai-runtime/__tests__ scripts/extensions/trend-finder/__tests__ scripts/extensions/trend-finder/sources/__tests__ src/lib/__tests__/trend-finder-schema.test.ts src/lib/__tests__/trend-finder-dashboard.test.tsx src/lib/__tests__/trend-finder-engine-replay.test.tsx src/lib/__tests__/trend-finder-collector.test.ts
bun run test
bun run test:e2e -- tests/e2e/trend-finder.spec.ts tests/e2e/trend-finder-engine-replay.spec.ts
bun run lint
bun run format:check
bun run build
bun run budget:check
bun run runtime:check-private
```

This pass is credential-free. Optional live Apify or Codex-account checks may be run from ignored local credentials, but submission claims must match the visible dashboard labels.

When running e2e certification with a fresh isolated `HOME`, install Chromium into that home first:

```bash
bunx playwright install chromium
```

### `typecheck`

Run the TypeScript compiler in check-only mode.

```
bun run typecheck
```

* **Prerequisite**: `bun install`
* **Side effects**: None (`--noEmit` check mode)
* **Expected output**: No output on success
* **Pass**: Exit code 0
* **Fail**: Exit code non-zero, type errors printed to stderr

### `lint`

Run ESLint across the project in check mode.

```
bun run lint
```

* **Prerequisite**: `bun install`
* **Side effects**: None
* **Expected output**: No output on success, or list of warnings/errors
* **Pass**: Exit code 0
* **Fail**: Exit code non-zero, lint errors printed

### `typecheck:scripts`

Run the script TypeScript project in check-only mode.

```
bun run typecheck:scripts
```

* **Prerequisite**: `bun install`
* **Side effects**: None
* **Expected output**: No output on success
* **Pass**: Exit code 0
* **Fail**: Exit code non-zero, script type errors printed

### `lint:fix`

Run ESLint with auto-fix enabled.

```
bun run lint:fix
```

* **Prerequisite**: `bun install`
* **Side effects**: Modifies files to fix auto-fixable lint issues
* **Expected output**: Summary of fixes applied
* **Pass**: Exit code 0
* **Fail**: Exit code non-zero, unfixable errors remain

### `format`

Auto-format all source files with Prettier.

```
bun run format
```

* **Prerequisite**: `bun install`
* **Side effects**: Writes formatted files in-place (`--write`)
* **Expected output**: List of files checked
* **Pass**: Exit code 0
* **Fail**: Prettier parse errors

### `format:check`

Validate formatting without modifying files. CI-safe.

```
bun run format:check
```

* **Prerequisite**: `bun install`
* **Side effects**: None (`--check` mode)
* **Expected output**: No output on success, or list of unformatted files
* **Pass**: Exit code 0 (all files formatted)
* **Fail**: Exit code non-zero (unformatted files listed)

### `budget:check`

Validate generated client bundle budgets after `bun run build`.

```
bun run budget:check
```

* **Prerequisite**: `bun run build`
* **Side effects**: None
* **Expected output**: Chunk and total client JS budget table
* **Pass**: Exit code 0, no budget violations
* **Fail**: Exit code non-zero, violating chunks listed

***

## Maintenance

### `clean`

Remove build artifacts.

```
bun run clean
```

* **Prerequisite**: None
* **Side effects**: Deletes `dist/`, `.output/`, `.vinxi/` directories
* **Expected output**: None
* **Pass**: Exit code 0, directories removed (or already absent)
* **Fail**: Permission errors

### `seed:data`

Copy the example data file if live-data.json is missing. Runs automatically before `dev` and `build`.

```
bun run seed:data
```

* **Prerequisite**: `src/data/live-data.example.json` exists
* **Side effects**: Creates `src/data/live-data.json` only if missing
* **Expected output**: None
* **Pass**: Exit code 0
* **Fail**: Example file missing

### `postinstall`

Re-activate the repository-local git hooks directory after dependency install. Runs automatically after `bun install`.

```
bun run postinstall
```

* **Prerequisite**: Git repo with `.githooks/` present
* **Side effects**: Sets local `core.hooksPath` to `.githooks`
* **Expected output**: None
* **Pass**: Exit code 0, including graceful no-op outside git or without `.githooks/`
* **Fail**: Git config write error

***

## Project-Specific

### `setup`

One-time machine setup. It scans local data, installs inherited starter skill assets, wires git hooks, prints scheduled aggregate guidance, and runs the local aggregator.

```
bun run setup
```

* **Prerequisite**: First run on a new machine
* **Side effects**: Populates `src/data/live-data.json`, installs inherited local skills, wires `.githooks/pre-commit`, and prints scheduled aggregate guidance.
* **Pass**: Setup completes without errors
* **Fail**: Any required setup step fails. Setup records failed steps, prints `Setup completed with failures.`, and exits nonzero instead of printing final `Done`.
* **Scheduled aggregate note**: Scheduler metadata defaults to enabled, but setup does not install or manage an aggregate timer.

### `aggregate`

Re-scan local data sources, run enabled extension collectors, and regenerate live-data.json.

```
bun run aggregate
```

* **Prerequisite**: API keys in `.env.local` (optional)
* **Side effects**: Overwrites `src/data/live-data.json` and writes private JSONL traces to `logs/aggregate-*.jsonl` plus `logs/aggregate.latest.jsonl`
* **Pass**: Exit code 0, data file and latest aggregate trace updated
* **Scheduler parity**: This remains the direct refresh path. The scheduler path delegates to the same aggregate implementation instead of duplicating collection behavior.

### `trend-finder:backtest`

Run the script-only Trend Finder historical backtest CLI.

```
bun run trend-finder:backtest -- --sources <source-id> --window-start <iso> --window-end <iso>
```

* **Prerequisite**: `bun install`
* **Side effects**: Writes private backtest archives under `.cache/extensions/trend-finder/backtests/`
* **Pass**: Exit code 0, private archives written, and optional bounded browser summary published only when `--publish-aggregate` is supplied
* **Fail**: Invalid source allowlist, window bounds, output path, or leakage guard violation
* **Notes**: Use `--dry-run` to validate the plan without writing backtest files.

### `trend-finder:export-brief`

Export the current generated Trend Finder Brief as a local standalone HTML report.

```
bun run trend-finder:export-brief -- --input <path> --output <dir> --dry-run --json
bun run trend-finder:export-brief -- --output <dir> --archive
```

* **Prerequisite**: A generated `src/data/live-data.json` containing `LiveData.extensions.items["trend-finder"].data` (run `bun run aggregate` or `bun run scheduler:trend-finder:run` first if it is missing)
* **Side effects**: Without `--archive`, writes `index.html` and `manifest.json` under `.cache/extensions/trend-finder/static-brief/` (override with `--output`). With `--archive`, writes `run-NNNN/index.html`, `run-NNNN/manifest.json`, optional run-local assets, and `latest.html` under the selected output directory.
* **Expected output**: Export summary with the resolved output directory, written file list, QA state, and archive run/latest pointer fields when archive mode is enabled; `--json` prints the machine-readable result
* **Pass**: Exit code 0 with default `index.html` and `manifest.json` written, an archived numbered run plus `latest.html` written, or a validated plan with no writes when `--dry-run` is supplied
* **Fail**: Missing live data, missing Trend Finder extension data, invalid Trend Finder payload, required-derived-field omissions, path traversal, unsafe output root, static Brief QA failure, private-string leak, media-embed output, unsafe latest pointer, or write failure
* **Notes**: Reads `src/data/live-data.json` by default; `--input` selects a different generated payload. Archive mode is local-only generated output for comparing human-readable Briefs across runs. Generated reports remain browser-safe local artifacts; do not commit `.cache/` outputs.

### `trend-finder:replay-backhistory`

Replay private Trend Finder backhistory caches into Phase 14 backtest artifacts.

```
bun run trend-finder:replay-backhistory -- --input-dir <path>
```

* **Prerequisite**: Private backhistory evidence cache under `.cache/extensions/trend-finder/backhistory/`
* **Side effects**: Writes private replay archives under `.cache/extensions/trend-finder/backtests/`; optionally writes bounded browser aggregate data with `--publish-aggregate`; optionally seeds normal private history caches with `--seed-current-history`
* **Expected output**: Replay summary with run ID, evidence/window/topic counts, archive path, and publish/seed status
* **Pass**: Exit code 0, archive and summary JSON written inside the backtest cache
* **Fail**: Missing cache files, invalid args, path escape guard, or write failure
* **Privacy**: Input evidence and generated archives are private runtime data. Do not commit `.cache/` outputs.

### `auth:openai`

Manage ignored OpenAI account auth used by optional Codex-account smoke checks.

```
bun run auth:openai -- status
bun run auth:openai -- login
```

* **Prerequisite**: Node runtime; browser login for `login`/`reauth`
* **Side effects**: `login`, `reauth`, `refresh`, and `logout` mutate the ignored local auth file; `status` only reports current metadata
* **Runtime behavior**: Codex-account AI runtime consumers attempt `refresh` automatically when stored auth is expired; rejected refresh tokens still require `reauth`
* **Expected output**: Redacted auth path, account status, expiry state, and next command hint; failures include a redacted diagnostic block with the attempted action, OAuth details when present, stored credential state, credential effect, interpretation, and next commands; `--json` emits a stable JSON form
* **Pass**: Exit code 0 for known commands that complete
* **Fail**: Unknown command, invalid args, invalid auth file, or failed login
* **Privacy**: Stored credentials, refresh tokens, and full account identifiers are not printed.

### `codex:smoke`

Run a minimal optional Codex-account text prompt through stored OpenAI account auth.

```
bun run codex:smoke -- --json
```

* **Prerequisite**: `bun run auth:openai -- login` or another valid stored auth file
* **Side effects**: Sends one network request to the Codex backend; no source files are modified
* **Expected output**: Redacted request/auth metadata, response status, usage, event count, and `secretValuesPrinted: false`
* **Pass**: Exit code 0, response text received
* **Expected missing-auth status**: Exit code 1 with `error.code: "missing_auth"`, redacted recovery commands, and `secretValuesPrinted: false`
* **Fail**: Missing/invalid stored auth, transport error, invalid args, or backend failure. Missing auth is expected in credential-free certification.
* **Privacy**: Auth contents and generated text are not printed in JSON mode; text output only reports whether text was received.

### `agents:codex:smoke`

Validate the planned Agents SDK Codex smoke command surface.

```
bun run agents:codex:smoke -- --json
```

* **Prerequisite**: None
* **Side effects**: None
* **Expected output**: Deferred-status payload with `agentsSdkImported: false` and `secretValuesPrinted: false`
* **Pass**: Help exits 0; the smoke surface emits the expected deferred payload
* **Expected deferred status**: Normal smoke execution exits 78 with `status: "deferred"` until `phase05-session04-ai-runtime-provider-boundary` implements the real Agents SDK provider check; invalid args exit 64
* **Privacy**: Does not import `@openai/agents`, read auth, send prompts, or print transcripts.

### `apify:smoke`

Check Trend Finder Apify configuration, readiness, and optional live actor validation.

```
bun run apify:smoke -- --json
bun run apify:smoke -- --json --tiny-validate
```

* **Prerequisite**: `APIFY_TOKEN` and explicit source config for live validation; default mode can run without credentials
* **Side effects**: Default mode reports metadata only. `--check-sources` and `--tiny-validate` may run Apify Actors and consume credits.
* **Expected output**: Redacted readiness, configured source counts, declaration summary, defaults, optional source/tiny-validation summaries, and `secretValuesPrinted: false`
* **Pass**: Exit code 0 for ready/setup-required/configured-with-warning states that do not have source-check failures
* **Expected missing-token status**: Exit code 0 with `status: "setup-required"` and `readiness.status: "missing-token"`. With `--tiny-validate`, live validation does not run and sources are blocked with `reason: "missing-token"`.
* **Fail**: Usage errors or live source/tiny-validation failure
* **Privacy**: Token values, Actor inputs, Dataset items, and response bodies are not printed.

### `scheduler:run`

Run the compatibility full aggregate scheduler job through the AI OS scheduler runner.

```
bun run scheduler:run
```

* **Prerequisite**: `bun install`; optional provider/source credentials remain in `.env.local` or the shell.
* **Side effects**: Runs the aggregate handler, updates `src/data/live-data.json`, writes sanitized private scheduler run state, writes a private scheduler log, and uses the aggregate scheduler lock.
* **Expected output**: `AI OS scheduler job "aggregate" completed.` on success.
* **Pass**: Exit code 0, latest private scheduler state records success, warning-only, degraded, or stale-lock recovery.
* **Fail**: Exit code non-zero for invalid args, unknown job, blocked run, timeout, unavailable job, or handler failure.
* **Privacy**: Private scheduler state/log files are local diagnostics. Command output does not print raw logs, command bodies, credentials, auth JSON, prompts, transcripts, source dumps, or private path contents.

### `scheduler:status`

Inspect the safe aggregate scheduler status surface and local timer intent.

```
bun run scheduler:status
```

* **Prerequisite**: `bun install`
* **Side effects**: None
* **Expected output**: Human-readable aggregate scheduler status with lifecycle, schedule, cadence, enablement stance, local config state, latest run label, warning count, timing metadata, lock outcome, generated-data freshness, command hints, and Dream boundary copy.
* **Pass**: Exit code 0, status prints without exposing private diagnostics.
* **Fail**: Exit code non-zero for invalid job IDs, unknown jobs, unsupported jobs, or unreadable status failures.
* **JSON form**: `bun run scripts/scheduler-status.ts --job aggregate --json`
* **No-gateway stance**: Scheduler status is command-only. The separate dashboard aggregate refresh uses local loopback middleware, not a scheduler gateway or remote API.

### `scheduler:agents:run`

Run the scoped host/local agent aggregate scheduler job.

```
bun run scheduler:agents:run
```

* **Prerequisite**: `bun install`; local source access as needed.
* **Side effects**: Refreshes host/local AI OS data, preserves `LiveData.extensions`, writes sanitized private scheduler state/logs, and uses the `agent-aggregate` scheduler lock plus the shared generated-data lock.
* **Expected output**: `AI OS scheduler job "agent-aggregate" completed.` on success.
* **Privacy**: Command output stays sanitized; private diagnostics remain local.

### `scheduler:agents:status`

Inspect the safe host/local agent aggregate scheduler status surface.

```
bun run scheduler:agents:status
```

* **Prerequisite**: `bun install`
* **Side effects**: None
* **Expected output**: Human-readable status for `agent-aggregate` with command hints, cadence intent, latest run state, warning count, and generated-data freshness.

### `scheduler:trend-finder:run`

Run the scoped Trend Finder scheduler job.

```
bun run scheduler:trend-finder:run
```

* **Prerequisite**: `bun install`; `VITE_CLAUDE_OS_ENABLED_EXTENSIONS=trend-finder` or equivalent shell enablement. Optional source/provider credentials remain in `.env.local` or the shell.
* **Side effects**: Refreshes only `LiveData.extensions.items["trend-finder"]`, preserves host/local branches and other extension items, writes sanitized private scheduler state/logs, and uses the `trend-finder` scheduler lock plus the shared generated-data lock.
* **Expected output**: `AI OS scheduler job "trend-finder" completed.` on success, or `skipped` when Trend Finder is not enabled.
* **Privacy**: Command output stays sanitized; raw source payloads, prompts, provider responses, tokens, and private paths are not printed.

### `scheduler:trend-finder:status`

Inspect the safe Trend Finder scheduler status surface.

```
bun run scheduler:trend-finder:status
```

* **Prerequisite**: `bun install`
* **Side effects**: None
* **Expected output**: Human-readable status for `trend-finder` with source and provider readiness represented as safe state, warning count, latest run state, generated-data freshness, and command hints.

### `scheduler:dream:run`

Run the AI OS Dream scheduler job through the AI OS scheduler runner.

```
bun run scheduler:dream:run
```

* **Prerequisite**: `bun install`; optional provider/source credentials remain in `.env.local` or the shell.
* **Side effects**: Runs the Dream handler, refreshes stale or missing host/local material through agent aggregate, writes private Dream output and continuity state, updates private scheduler run state, writes a private scheduler log, and uses the Dream scheduler lock.
* **Expected output**: `AI OS scheduler job "dream" completed.` on success.
* **Pass**: Exit code 0, latest private scheduler state records success, warning-only, degraded, skipped, stale-lock recovery, or other handled Dream states.
* **Fail**: Exit code non-zero for invalid args, unknown job, blocked run, timeout, unavailable job, or handler failure.
* **Privacy**: Private Dream output, continuity state, scheduler state, and logs are local diagnostics. Command output does not print raw logs, command bodies, credentials, auth JSON, prompts, transcripts, source dumps, or private path contents.

### `scheduler:dream:status`

Inspect the safe Dream scheduler status surface.

```
bun run scheduler:dream:status
```

* **Prerequisite**: `bun install`
* **Side effects**: None
* **Expected output**: Human-readable Dream scheduler status with lifecycle, schedule, cadence, enablement stance, latest run label, warning count, timing metadata, lock outcome, private Dream output label, and command hints.
* **Pass**: Exit code 0, status prints without exposing private diagnostics.
* **Fail**: Exit code non-zero for invalid job IDs, unknown jobs, unsupported jobs, or unreadable status failures.
* **JSON form**: `bun run scripts/scheduler-status.ts --job dream --json`
* **No-gateway stance**: Scheduler status is command-only. The separate dashboard Dream button uses local loopback middleware, not a scheduler gateway or remote API.

### `runtime:check-private`

Verify generated runtime, auth, cache, log, coverage, and browser-test artifacts are ignored and untracked using git metadata only.

```
bun run runtime:check-private
```

* **Prerequisite**: `git` available
* **Side effects**: None
* **Expected output**: Private runtime artifact check passed
* **Pass**: Exit code 0, private paths are ignored and untracked
* **Fail**: Exit code non-zero, tracked or unignored private artifact path listed

### `install-dream`

Install the daily AI OS Dream schedule. macOS uses launchd, Windows uses Task Scheduler, and Linux prints a manual crontab line.

```
bun run install-dream
```

* **Prerequisite**: Bun available to the OS scheduler path
* **Side effects**: Creates or replaces the launchd plist or Windows scheduled task; Linux prints guidance only
* **Pass**: Schedule installed, or Linux guidance printed

### `uninstall-dream`

Remove the daily AI OS Dream schedule on platforms managed by the installer.

```
bun run uninstall-dream
```

* **Side effects**: Removes launchd plist or Windows scheduled task; Linux requires manual crontab edits
* **Pass**: Schedule removed, or Linux manual-removal guidance printed

***

## Testing

### `test`

Run all Vitest tests once.

```
bun run test
```

* **Prerequisite**: `bun install`
* **Side effects**: None
* **Expected output**: Test summary with pass/fail counts
* **Pass**: Exit code 0, all tests pass
* **Fail**: Exit code non-zero, failing tests listed

### `test:watch`

Run Vitest in watch mode (re-runs on file changes).

```
bun run test:watch
```

* **Prerequisite**: `bun install`
* **Side effects**: None (long-running process)
* **Expected output**: Interactive Vitest watcher
* **Pass**: Tests re-run and pass on save
* **Fail**: Test failures shown inline

### `test:coverage`

Run all tests with coverage reporting.

```
bun run test:coverage
```

* **Prerequisite**: `bun install`
* **Side effects**: Writes coverage output
* **Expected output**: Test summary plus coverage table
* **Pass**: Exit code 0
* **Fail**: Exit code non-zero, failing tests or coverage threshold violations

### `test:e2e`

Run browser-visible Playwright coverage for AI OS and Trend Finder routes.

```
bun run test:e2e
```

* **Prerequisite**: Playwright browser installed; `bun install` completed
* **Side effects**: Starts a local Vite web server on `127.0.0.1:5189` unless `PLAYWRIGHT_REUSE_EXISTING_SERVER=true`; seeds `src/data/live-data.json` if missing; writes Playwright HTML/report artifacts on failure or completion
* **Expected output**: Playwright project summary for Chromium
* **Pass**: Exit code 0, all e2e specs pass
* **Fail**: Route load timeout, assertion failure, server startup failure, or browser dependency issue

***

## Quality Gate Sequence

Run all quality checks in order for a full validation pass:

```bash
bun install
bun run typecheck
bun run typecheck:scripts
bun run format:check
bun run lint
bun run test
bun run test:e2e -- tests/e2e/trend-finder.spec.ts tests/e2e/trend-finder-engine-replay.spec.ts
bun run build
bun run budget:check
bun run runtime:check-private
```

All commands must exit 0 for the gate to pass.


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter, and the optional `goal` query parameter:

```
GET https://ai-os-and-trend-finder.gitbook.io/ai-os-and-trend-finder-docs/docs/commands.md?ask=<question>&goal=<endgoal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.