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/yc-software/qm/interactive-loginnpx skills add yc-software/qm --skill interactive-logingit clone --depth 1 https://github.com/yc-software/qmWhat 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 | $0.00046 | $0.01196 |
| Opus 5 | $0.00023 | $0.00598 |
| Sonnet 5 | $0.00009 | $0.00239 |
| Haiku 4.5 | $0.00005 | $0.00120 |
Grade A, and why
interactive-login 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 2d 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 — 85 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Interactive login (and what backgrounding can and can't do)
Some logins are interactive: the CLI prints a verification URL + one-time code, then
polls the provider until the human approves in a browser, and only then writes the
token to $HOME. Run that as a plain blocking execute and it either hangs your whole
turn waiting for a human, or — worse — the turn's sandbox is torn down at turn end and
the poller dies before the human finishes, so nothing is ever saved. The next turn
starts over and loops forever.
You usually don't have to manage any of this. Just run the native login command.
aws sso login --use-device-code
AWS (and CLIs that log in through AWS SSO): always pass
--use-device-code(aws sso login≥ CLI v2 2.22). The bare form uses a PKCE flow that redirects to127.0.0.1and only completes if the approving browser is on the same machine as the CLI. Since the user approves on their computer, that flow can never finish on the agent computer;--use-device-codeforces the device-grant (URL + code, completed by server-side polling) that works cross-device.
On the resident, per-scope agent computer (the production substrate), the platform
recognizes these as interactive logins and runs each as a durable process session
(ADR 0002): it returns the verification URL + code to you immediately instead of
blocking, and keeps polling the provider on the agent computer across turns. (This
recognition is wired only where the backend supports process sessions — on a per-turn
local/docker sandbox nothing intercepts the command and it just blocks; see "When it
will NOT work".) When the human approves, the poller writes the token into the
durable $HOME and exits. Give the user the URL + code, tell them to approve, then on
a later turn re-run the same command (or any command for that tool) — if approval
landed, the platform reports you're already authenticated; if not, you get the same URL
- code back. Each pending login self-expires (~10–15 min); if it lapses, just run the login again.
So the right pattern is: run the native login, hand over the URL + code, continue, and re-check later. You do not poll in a loop, you do not hold the turn open, and you do not hand-roll the exchange yourself.
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.
- 2d ago First seen · 85 lines · 46 tokens per session scan A 4adc90342ddc
interactive-login is a skill published in the GitHub repository yc-software/qm (14,428 stars, last pushed today), licensed MIT. It adds 46 tokens to every session and 1,196 once invoked, about $0.0002 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
vercel-sandbox
Run agent-browser + Chrome inside Vercel Sandbox microVMs for browser automation from any Vercel-deployed app. Use when the user needs browser automation in a Vercel app (Next.js, SvelteKit, Nuxt, Remix, Astro, etc.), wants to run headless Chrome without binary size limits, needs persistent browser sessions across…
protected-vercel-deployments
Access and test Vercel deployments protected by Vercel Authentication, SSO, or Deployment Protection with agent-browser. Use when a preview or production URL redirects to a Vercel login page, returns a protection 401 or 403, or needs short-lived Trusted Sources OIDC authentication instead of a static bypass secret or…
firecrawl
Search the web and scrape pages into clean markdown with the Firecrawl API — query-based discovery, single-URL extraction including public PDFs, driven by curl with a vault-stored API key.
penguin-harness-manual-test
Use when standing PenguinHarness up to try a change by hand — launching the Web App, the desktop shell, the landing page or the docs site to click through it, screenshot it, or reproduce a report. Covers the four dev entry points and their ports, which data root each writes to, and the four ways a healthy setup looks…
web-design
网页设计与部署。生成精美的单页 HTML 网页（报告、落地页、数据可视化等），支持一键部署到 Cloudflare Pages。使用 Tailwind CSS + Chart.js + Font Awesome 技术栈。当用户要求制作网页、生成报告页面、创建落地页、数据可视化展示、部署网页到线上时使用。.
vercel-cli
Deploy apps to Vercel. Use when asked to deploy, ship, or publish a web application, or manage Vercel projects, domains, and environment variables.