---
name: developing-with-streamlit-in-snowflake
description: "Use for Streamlit development tasks with a Snowflake angle: Snowflake-connected dashboards, Streamlit-in-Snowflake (SiS) deployment to warehouse / SPCS / Workspaces, applying Snowflake branding, st.connection('snowflake'), troubleshooting a local `streamlit run` against Snowflake (wrong role/user/database, 'Database not authorized', PAT-bound USE ROLE failure, stale st.connection cache), and operating an already-deployed STREAMLIT object (ALTER STREAMLIT SET QUERY_WAREHOUSE, RENAME, DROP, GRANT, SHOW STREAMLITS). Also use for general Streamlit authoring (widgets, layouts, caching, theming, custom components) — this skill routes general OSS questions to version-matched content from a detected Streamlit ≥1.57 install, or to a bundled OSS snapshot when no install is available. Triggers: streamlit, st., dashboard, app.py, theme, beautify, style, CSS, color, background, button, custom component, st.components, snowflake dashboard, monitor snowflake, streamlit on snowflake, streamlit in snowflake, SiS, scaffold, snowflake theme, st.connection snowflake, snow streamlit deploy, deploy this streamlit, redeploy, alter streamlit, show streamlits, drop streamlit, rename streamlit app, change query warehouse, streamlit app down, streamlit run wrong role, database not authorized, SNOWFLAKE_DEFAULT_CONNECTION_NAME."
---

# Developing with Streamlit in Snowflake

Entry-point skill for Streamlit work in a Snowflake context. Routes to one of two specialized sub-skills based on the user's prompt.

## Sub-skills

| Sub-skill | Read when |
|---|---|
| `sf/SKILL.md` (skill name: `scaffolding-streamlit-in-snowflake`) | User wants a Snowflake-wired starter (dashboard, theme, SiS deploy) or any Snowflake-specific Streamlit task |
| `developing-with-streamlit/SKILL.md` | General Streamlit authoring with no Snowflake-specific angle (used as fallback when no Streamlit ≥1.57 install is detected) |

Execute the steps below in order. Do not answer the user's Streamlit question until a guidance source is loaded.

## Step 1 — Identify routing

If the user's prompt involves any of:
- Snowflake-connected data, `st.connection("snowflake")`
- Deploying to Streamlit-in-Snowflake (SiS warehouse, SPCS, Workspaces), `snow streamlit deploy`, `snowflake.yml`
- Snowflake-branded theming
- A Snowflake-specific scaffold (compute monitor, metrics, stock peers, etc.)
- **Troubleshooting a local `streamlit run` that talks to Snowflake** — wrong role / user / database in the running session, "Database X does not exist or not authorized" while user has Snowsight access, PAT-bound `USE ROLE` failures, stale `st.connection` cache after env-var change, `SNOWFLAKE_DEFAULT_CONNECTION_NAME` questions
- **Operating an already-deployed `STREAMLIT` object** — `ALTER STREAMLIT … SET QUERY_WAREHOUSE`, `RENAME TO`, `DROP STREAMLIT`, `GRANT USAGE`, `SHOW STREAMLITS`, "change the warehouse my deployed app uses", "rename my deployed app", "where are the logs for my SiS app" (there aren't any)

→ Read `<SKILL_DIR>/sf/SKILL.md` and follow its guidance (including `sf/references/snowflake-deployment.md` for manifests, `compute_pool`, and deploy). Continue to Step 2.

Otherwise (general Streamlit authoring with no Snowflake angle) → continue to Step 2.

## Step 2 — OSS path: locate Streamlit content

### 2a. Detect the Python environment

Use the built-in `cortex env detect` command, passing the user's project directory as `--dir` (absolute path):

```bash
cortex env detect --dir <absolute path to user's project>
```

The command returns JSON:

```json
{"directory": "...", "result": "..."}
```

When environments are found, the `result` string embeds a JSON array with an entry per environment:

```json
[{"dir": "/abs/path/to/project", "cmd": "uv run python"}, ...]
```

Each entry gives the env's directory (`dir`) and the exact command to invoke its Python (`cmd`). If `result` reports that no environments were found, skip to **Case B**.

### 2b. Probe each environment for Streamlit

For each entry, invoke the `cmd` against the one-line probe below. For `uv run` style invocations, pin the project with `--project <dir>` so cwd does not matter; for path-based invocations (e.g. `.venv/bin/python`), form an absolute path with the entry's `dir`.

Probe body:

```python
import streamlit, os
print(f"STREAMLIT_PATH={os.path.dirname(streamlit.__file__)}")
print(f"STREAMLIT_VERSION={streamlit.__version__}")
```

Concretely, for a `uv run python` entry:

```bash
uv run --project <dir> python -c "import streamlit, os; print(f'STREAMLIT_PATH={os.path.dirname(streamlit.__file__)}'); print(f'STREAMLIT_VERSION={streamlit.__version__}')"
```

Capture `STREAMLIT_PATH` and `STREAMLIT_VERSION` from the first environment where the probe exits 0. If every environment's probe fails, skip to **Case B**.

## Step 3 — Delegate based on detection outcome

### Case A — Streamlit detected AND version ≥ 1.57.0

The installed package ships version-matched skill content. Read and follow:

```
<STREAMLIT_PATH>/.agents/skills/developing-with-streamlit/SKILL.md
```

Treat that file as your authoritative guidance source for the rest of the task. Reference files, templates, and assets it points to live under `<STREAMLIT_PATH>/.agents/skills/developing-with-streamlit/` — load them from there, not from this skill's directory.

### Case B — No Streamlit found OR version < 1.57.0

Fall back to the bundled OSS snapshot under this skill's `developing-with-streamlit/` sub-skill — content synced from the latest Streamlit PyPI wheel (`.synced-from-version` records the exact version). Read and follow:

```
<SKILL_DIR>/developing-with-streamlit/SKILL.md
```

Reference files and templates it points to live under `<SKILL_DIR>/developing-with-streamlit/references/` and `<SKILL_DIR>/developing-with-streamlit/assets/`.

**Version caveat**: the bundled snapshot is not version-matched to the user's installed Streamlit (if any). If a suggestion you make fails on import or at runtime, ask the user which Streamlit version they're on and adjust.

## Resources

- Streamlit API reference: https://docs.streamlit.io/develop/api-reference