Agent skill · Audited from source

Vellum Oauth Integrations

skill · vellum-ai/vellum-assistantMIT licenceAudited 14 Oct 2026 · source 19bc86343b63

★ 1,425pushed 1 day ago
✓SAFEGrade B · 89/100 · 14 Oct 2026 · 19bc863
1,425repo stars · whole vellum-ai/vellum-assistant

Is it safe? Yes — SAFE, grade B (89/100). We read this skill's source at commit 19bc86343b63 and found no filesystem, network or shell access and no hard-coded secrets. What it does: Act on behalf of your user in any third-party software that supports OAuth 2.0

Ask an AI about this pageGChatGPTCClaudePPerplexityGeGeminiXGrok
More
Cite

Cite this page

ModelsAtlas. “Vellum Oauth Integrations (vellum-ai/vellum-assistant)”. Retrieved 14 Oct 2026. https://modelsatlas.com/skills/vellum-ai/vellum-oauth-integrations

Trust audit

Audit 0.4.1 · source read 14 Oct 2026

Trust score89/100✓SAFE B
0 critical0 high0 medium6 low6 info

SAFE means nothing in the source contradicts what the skill says it does. Grade B is shared by 13,985 of 14,816 audited skills.

What the source does

  • FilesystemNone observedNo reads or writes in the source
  • NetworkNone observedNo outbound calls in the source
  • ShellNone observedNo commands run in the source
  • DependenciesPinnedNothing floats
  • SecretsNone foundNo keys in source

Five audit layers

LayerWhat we checkResultFindings
L0 ProvenanceRepo owner, licence, archive status, activityPASS5
L1 Static sourceSecrets, path traversal, supply chain, hidden codeNA0
L2 InstructionsPrompt injection, authority framing, hidden directivesPASS7
L3 InventoryThe files and declarations match the sourcePASS0
L4 SandboxRun in isolation and watch behaviourSKIPPED why0

Gates applied: no behavioural pass.

Findings (12)

LOWPrompt injectionprompt.transfer_instruction · references/provider-app-setups/airtable.md:11
F-01 Prompt injectionCWE-94 · CWE-1427Layer L2
- **Token endpoint auth method:** secret via POST body
Why
an instruction to move sensitive data to an outside destination
Fix
remove; a skill never needs the user's secrets off the machine
INFOPrompt injectionprompt.authority_framing · references/provider-app-setups/github.md:93
F-01 Prompt injectionCWE-94 · CWE-1427Layer L2
- `repo` - full access to repositories
Why
No further detail recorded for this rule.
Fix
—
INFOPrompt injectionprompt.credential_read · references/CONFIGURING_APPLICATIONS.md:48
F-01 Prompt injectionCWE-94 · CWE-1427Layer L2
Do NOT collect the client secret conversationally. Never solicit the secret in chat and never store a chat-pasted value with `assistant credentials set`. Always collect it through the secure `assistan
Why
asks the agent to read credentials
Fix
—
INFOPrompt injectionprompt.credential_read · references/provider-app-setups/airtable.md:88
F-01 Prompt injectionCWE-94 · CWE-1427Layer L2
Follow the `vellum-oauth-integrations` workflow to collect credentials, register the OAuth app, connect, and verify.
Why
asks the agent to read credentials
Fix
—
INFOPrompt injectionprompt.credential_read · references/provider-app-setups/asana.md:72
F-01 Prompt injectionCWE-94 · CWE-1427Layer L2
Follow the `vellum-oauth-integrations` workflow to collect credentials, register the OAuth app, connect, and verify.
Why
asks the agent to read credentials
Fix
—
INFOPrompt injectionprompt.credential_read · references/provider-app-setups/discord.md:97
F-01 Prompt injectionCWE-94 · CWE-1427Layer L2
Follow the `vellum-oauth-integrations` workflow to collect credentials, register the OAuth app, connect, and verify.
Why
asks the agent to read credentials
Fix
—
INFOPrompt injectionprompt.credential_read · references/provider-app-setups/dropbox.md:108
F-01 Prompt injectionCWE-94 · CWE-1427Layer L2
Follow the `vellum-oauth-integrations` workflow to collect credentials, register the OAuth app, connect, and verify.
Why
asks the agent to read credentials
Fix
—

Audit 0.4.1 · source sha 19bc86343b63 · 14 Oct 2026Report an issue or request a re-scan

Provenance

Who published it, and what the verdict is pinned to

  • Publishervellum-ai GitHub organizationOwner of the repository this skill is read from
  • Source foldervellum-ai/vellum-assistant/skills/vellum-oauth-integrationsThe exact folder we audited
  • LicenceMITAs the repository states it
  • Commit pinned for this verdict19bc86343b63A new audit runs when the source changes
  • Repository statusNot archivedLast push 1 day ago
  • Sandboxed runNot runSkills are read, not executed; layer L4 is skipped

What it tells the agent

Full source on GitHub ↗

Its description: Act on behalf of your user in any third-party software that supports OAuth 2.0

---
name: vellum-oauth-integrations
description: Act on behalf of your user in any third-party software that supports OAuth 2.0
compatibility: "Designed for Vellum personal assistants"
metadata:
  emoji: "🔌"
  vellum:
    category: "integrations"
    display-name: "Vellum OAuth Integrations"
---

## Using OAuth Integrations

Integrating with a third-party software via OAuth is typically used to perform actions on behalf of the user. It involves having the user log in using their own credentials, specifying the scopes that they want to provide, and then afterwards, API requests can be made on their behalf.

**Important:** Avoid reaching for an OAuth integration if your intent is to act as yourself rather than as your user.

## The Assistant OAuth CLI

You are provided with the `assistant oauth` CLI for performing all necessary oauth-related actions.

**Important:** When in doubt how a command works or how to do something, read the references at the bottom. Never guess at how the CLI works. Read references and use the `--help` flag for any command you're about to run.

## Viewing Available OAuth Providers

Vellum assistants can natively integrate with any application that supports OAuth 2.0. Many OAuth providers come pre-configured and ready to use. You can view them by running:

```bash
assistant oauth providers list
```

You can also search for specific providers. Here's an example that searches for the "google" provider:

```bash
assistant oauth providers list --provider-key google
```

## Managed vs Your-Own Mode

All providers support "your-own" mode and some support "managed" mode.

### Default: Managed Mode

"managed" mode relies on a first-class integration with the Vellum Platform. Managed mode is typically easier to set up and get going, often only requiring the user to log in with no additional configuration needed before they can begin using the integration. Managed mode is the recommended method for most users, especially those that are less technical or newer to their Vellum assistant.

When a task needs a managed OAuth connection, present it in chat with `ui_show` using `surface_type: "oauth_connect"` instead of sending the user to Settings or trying to open the OAuth flow from shell. See [Connecting Accounts](references/CONNECTING_ACCOUNTS.md) for the exact surface shape.

Note that using managed mode:

- Requires an account with the Vellum Platform
- May result in billable usage
- Requires that requests to the third party are sent through Vellum's servers

### Your-Own Mode

"your-own" mode requires that the user creates their own OAuth application directly with the third-party and then enter the application's Client ID and Client Secret into Vellum.

Your-own mode is typically best if:

- Vellum does not have a first-party integration with the provider and managed mode is not supported
- The user is more tech-savvy and comfortable setting up OAuth apps
- The user does not want to create an account with the Vellum Platform
- The user is more sensitive to potential billing implications
- The user is sensitive to their data going to the Vellum Platform

### How to Choose

**Default to managed mode whenever the provider supports it.** Check support with:

```bash
assistant oauth providers get <provider-key> --json | jq -r '.managedServiceConfigKey'
```

If this returns a non-null value, managed mode is supported and should be your starting recommendation. Only fall back to your-own mode if:

- The provider does not support managed mode (`managedServiceConfigKey` is null), OR
- The user explicitly wants their own OAuth app (e.g. they're running self-hosted, they're privacy-sensitive, they're configuring a per-assistant Slack app), OR
- The user already has a working your-own connection — don't re-recommend managed mid-flow for already-connected providers.

Do not present managed and your-own as equivalent choices to a fresh user. Lead with managed, then offer your-own as an alternative if they push back.

### Differentiating & Setting a Provider's Mode

You can determine whether a given provider supports managed mode based on the details returned by:

```bash
# Find the provider of interest in the list response
assistant oauth providers list

# Or, return just the provider of interest with their details
assistant oauth providers get <provider-key>
```

You can determine what mode the provider is currently set to use with:

```bash
assistant oauth mode <provider-key>
```

You can update which mode a given provider should use with:

```bash
assistant oauth mode <provider-key> --set "managed"|"your-own"
```

## Troubleshooting

### Permission / Insufficient Scope Errors

If an API request returns a `403 Forbidden`, `401 Unauthorized`, or a message about missing scopes, the connection likely doesn't have the scopes needed for that action. See [Updating Scopes](references/UPDATING_SCOPES.md) for how to disconnect and reconnect with the required scopes.

# Reference

For detailed information on the following topics, see the reference files:

- **[Registering New OAuth Providers](references/REGISTERING_PROVIDERS.md)** - How to register a new OAuth provider that doesn't come with Vellum's defaults
- **[Configuring a New OAuth Application](references/CONFIGURING_APPLICATIONS.md)** - How to configure a user-managed OAuth Application for providers whose mode is set to "your-own"
- **[Connecting Accounts](references/CONNECTING_ACCOUNTS.md)** - How to direct the user to log in and create a new OAuth connection
- **[Updating Scopes](references/UPDATING_SCOPES.md)** - How to update the scopes on an existing connection when additional permissions are needed
- **[Making Requests on Behalf of the User](references/MAKING_REQUESTS.md)** - How to make requests and take actions on behalf of the user once they have connected their account

Excerpt, first 5,837 characters of SKILL.md, verbatim from the audited commit. It is text from someone else’s repository, quoted so you can judge it before your agent reads it.

Install

Pinned to the audited commit 19bc863

Download folder ↗
git clone https://github.com/vellum-ai/vellum-assistant
git -C vellum-assistant checkout 19bc86343b63

The author documents no install command, so this uses the repository directly. Our verdict covers commit 19bc86343b63 only.

Works in

The author has not said which agents this skill supports, and we have not tested it in any.

Audit history

Every audit, every source change

  1. ✓SAFE B 89 / 100

    first audit · source 19bc86343b63

README badge

For maintainers · shows the latest audit

ModelsAtlas: SAFE BMODELSATLASSAFE · B
[![ModelsAtlas: SAFE B](https://modelsatlas.com/skills/vellum-ai/vellum-oauth-integrations/badge.svg)](https://modelsatlas.com/skills/vellum-ai/vellum-oauth-integrations/)

The badge shows our verdict and grade for this skill, and links back to this page. It changes when the audit does.

Add this badge to your README

[![ModelsAtlas: SAFE B](https://modelsatlas.com/skills/vellum-ai/vellum-oauth-integrations/badge.svg)](https://modelsatlas.com/skills/vellum-ai/vellum-oauth-integrations/)

Questions about Vellum Oauth Integrations

What does the Vellum Oauth Integrations skill do?
An AI Assistant that’s easy to setup, does your work 24/7, knows your preferences and gets better over time.
Is Vellum Oauth Integrations safe to install?
The audit found nothing in the source that contradicts what it says it does, and graded it B (89/100). Grade A is held back for packages that have also passed a sandboxed behavioural run, which is why a clean skill reads B.
What can Vellum Oauth Integrations access on my machine?
The audit observed no filesystem, network or shell use at all in its source.
What do I need installed to use Vellum Oauth Integrations?
Its own instructions reference ui_show. Dependencies are pinned to exact versions.
How current is this page?
The grade is for one exact copy of the source (19bc86343b63), read on 2026-10-14. The repository is watched, and a new audit runs when it changes — this is the first audit.

Sources

Spotted something wrong? Report a correction · Methodology

Keep going in your AI

Opens with this page as the source. We add nothing else to the prompt.