Vellum Assistant is a personal AI assistant that remembers information about users, learns their preferences, and takes actions across connected apps. It is intended for people who want an assistant that can manage conversations, unfinished work, and proactive notifications over time. The catalogue skills, hooks, instruction, and setting configure or extend how the assistant works.
Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add skills/vellum-ai/vellum-assistant/cli-testingnpx skills add vellum-ai/vellum-assistant --skill cli-testinggit clone --depth 1 https://github.com/vellum-ai/vellum-assistantWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/vellum-ai/vellum-assistant/cli-testing)<a href="https://agentmods.dev/skills/vellum-ai/vellum-assistant/cli-testing"><img src="https://agentmods.dev/badge/skills/vellum-ai/vellum-assistant/cli-testing.svg" alt="Measured on agentmods" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00066 | $0.01592 |
| Opus 5 | $0.00033 | $0.00796 |
| Sonnet 5 | $0.00013 | $0.00318 |
| Haiku 4.5 | $0.00007 | $0.00159 |
Grade A, and why
cli-testing scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 5d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 139 lines — stays where its author put it; the contents beside it link to each section on GitHub.
CLI Testing — Exercise the Assistant End-to-End
Drive a real assistant from the terminal only. The vellum CLI (cli/, package
@vellumai/cli) manages instance lifecycle; vellum message / vellum events
exercise a running instance. See cli/AGENTS.md and
the root README.md § CLI for command reference.
0. Prerequisites
export PATH="$HOME/.bun/bin:$PATH" # bun + the linked `vellum` binary
vellum ps # sanity check the CLI resolves
If vellum is missing, run ./setup.sh from the repo root once (installs deps,
links the vellum command). Docker must be running for the default flow below.
1. Provide an LLM provider key (from the environment)
Local-mode and Docker-mode instances need one LLM provider key. The CLI reads it straight from the host environment — just export it before hatching/setup:
export ANTHROPIC_API_KEY=sk-ant-... # or OPENAI_API_KEY / GEMINI_API_KEY /
# FIREWORKS_API_KEY / OPENROUTER_API_KEY /
# MINIMAX_API_KEY
In Devin sessions ANTHROPIC_API_KEY is typically already present in the
environment — check with echo "${ANTHROPIC_API_KEY:0:7}" before asking for one.
The CLI maps providers to env vars in
cli/src/shared/provider-env-vars.ts.
2. Hatch — default to a Docker hatch built from source
Always default to --remote docker. It runs the assistant, gateway, and
credential-executor in isolated containers that mirror production and keep the
test off your host process table. Reserve --remote local (§5) for the rare
case where Docker is unavailable.
Build from source — that's the point of testing. A bare
vellum hatch --remote docker pulls the published platform images even when
the CLI itself runs from your checkout, so it would test released code, not your
changes. Source-build is opt-in via a flag
(resolveDockerHatchMode in cli/src/lib/docker.ts):
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 5d ago First seen · 139 lines · 66 tokens per session scan A 143c254822b0
cli-testing is a skill published in the GitHub repository vellum-ai/vellum-assistant (1,193 stars, last pushed today), licensed MIT. It adds 66 tokens to every session and 1,592 once invoked, about $0.0003 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
playwright-pom-discovery
Use when building or extending a Page Object Model (POM) for the Opik E2E suite (under testsendtoend/e2e/pom/) and you need to choose stable selectors against the live UI. Walks through seeding required state, exploring the running page with the Playwright MCP (accessibility snapshot + data-testid enumeration)…
explore-feature
Use when a developer wants an e2e test covering a change they just made — e.g. "explore this feature", "add a test for my PR", "cover the feature in PR.
writing-e2e-tests
Use when a developer wants to add, write, or create an end-to-end test for an Opik feature, page, or branch — e.g. "add an e2e test for the experiments comparison page", "write a test for the feature I just built", "e2e test for this branch", "cover the dataset items flow with a test". Runs the full loop in…
debugging-e2e-tests
Use when an Opik E2E test has failed and a developer wants it investigated — e.g. "why did this e2e test fail?", "investigate the failing run on my PR", "is dataset-crud-smoke flaky?", "the nightly e2e suite went red". Takes a failure from a CI check, a TestOps launch, a test name, or a local run; gathers the trace…
writing-visual-tests
Use when a developer wants to add a visual regression (screenshot) test for an Opik UI page or panel — e.g. "add a visual test for the trace sidebar", "screenshot each tab of the dataset panel", "visual regression test for the new empty state". Covers the page-object pattern, seeding via the test-helper-service…
codex-app-parity
Use only when the user explicitly mentions Codex parity, codex-app-parity, Codex.app parity, or asks to compare against the installed Codex desktop app.