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 skills add HybridAIOne/hybridclaw --skill handoff-linksgit clone --depth 1 https://github.com/HybridAIOne/hybridclawWrote 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/hybridaione/hybridclaw/handoff-links)<a href="https://agentmods.dev/skills/hybridaione/hybridclaw/handoff-links"><img src="https://agentmods.dev/badge/skills/hybridaione/hybridclaw/handoff-links/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/hybridaione/hybridclaw/handoff-links"><img src="https://agentmods.dev/badge/skills/hybridaione/hybridclaw/handoff-links.svg" alt="Reviewed on agentmods" width="80" 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 5.1 | $0.00082 | $0.01747 |
| Opus 5.5 | $0.00033 | $0.00699 |
| Sonnet 5.5 | $0.00016 | $0.00349 |
| Haiku 4.5 | $0.00008 | $0.00175 |
Grade A, and why
handoff-links 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 12d 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.
Handoff Links
You can't order or book for the user at Lieferando, Wolt, Uber Eats, Knuspr, Gurkerl or Deutsche Bahn. Do the searching and thinking, then hand the last step to the user: a link that opens their app, or the website when the app is missing, with the search or the shopping list already filled in. The user picks, pays and confirms there.
Rules
- Never say you ordered, booked or reserved anything. Say what the link opens.
- Do not order, book or pay on these sites with the browser. Wolt's terms forbid bots, and the others block automated browsers. Web search to find options is fine.
- Build links only from the templates below. A restaurant page is fine only when a search result gave you its exact URL. Never invent a slug or an ID.
- Percent-encode every value: space
%20,ä%C3%A4,ö%C3%B6,ü%C3%BC,ß%C3%9F,(%28,)%29. A raw space breaks the link. - Write each link as a Markdown link on its own line. The label is short, in
the user's language, and says what opens, for example
[Sushi in 80331 bei Lieferando](https://...). Never print a bare URL. - One link per service and at most three in a reply.
- Use the postcode, city, home station and preferences you already know from memory or the conversation. When the link needs something you don't know, such as the postcode for Lieferando or the destination of a trip, ask once, briefly.
- Remember which delivery service the user chooses and offer it first next time.
Food delivery
Offer the user's usual service. If you don't know it, start with Lieferando, which covers the most of Germany and Austria, and add Wolt and Uber Eats in large cities.
| Service | Link |
|---|---|
| Lieferando, Germany | https://www.lieferando.de/lieferservice/essen/<postcode>?q=<dish or restaurant> |
| Lieferando, Austria | https://www.lieferando.at/lieferservice/essen/<postcode>?q=<dish or restaurant> |
| Wolt | https://wolt.com/de/<country>/<city>/search?q=<dish or restaurant> |
| Uber Eats | https://www.ubereats.com/de/search?q=<dish or restaurant> |
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.
- 12d ago Changed · -14 lines · -12 tokens per session f12d10365289
- 14d ago First seen · 153 lines · 94 tokens per session scan A 25edf1f0c71b
handoff-links is a skill published in the GitHub repository HybridAIOne/hybridclaw (159 stars, last pushed yesterday), licensed MIT. It adds 82 tokens to every session and 1,747 once invoked, about $0.0003 per session on Opus 5.5. 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 2026-10-03.
Other skills, from other repositories
apify-product-data-setup
Wire an AI agent to live e-commerce product data using Apify's E-commerce Scraping Tool over MCP, either as runtime tool calls or as a scheduled refresh into a vector store. Trigger on "give my agent live product data", "my agent quotes stale prices", "connect Apify MCP to Claude or Cursor or n8n", "add product data…
browser-f8f2966e
A browser Payment Request API pattern for checking whether a payment method can be used, starting a payment, and reading its status.
customer-360
A customer-profile workflow that combines customer records, business relationships, purchase history, contacts, opportunities, contracts, and orders into a structured overview. CRM means customer relationship management: records used to track business interactions.
lilbee-mcp
Search and manage the user's local lilbee knowledge base over MCP. Use whenever the user has indexed code, docs, PDFs, or web pages into lilbee and you need cited answers, or whenever they ask you to ingest content, swap models, or tune retrieval against their library. Every fact returned cites file and line.…
renewal-scan
A CRM scanning workflow for finding contracts nearing expiry and sales opportunities that have not been followed up. It reads contract and opportunity records and applies stated date rules.
lilbee-mcp-wiki
Wiki layer for lilbee. Use only when the user explicitly asks about wiki / concept / entity / synthesis pages, or when lilbeestatus shows a built wiki. Requires the lilbee-mcp skill to be active for the underlying MCP connection.