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 OpenLinkSoftware/ai-agent-skills --skill ucp-clientgit clone --depth 1 https://github.com/OpenLinkSoftware/ai-agent-skillsWrote 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/openlinksoftware/ai-agent-skills/ucp-client)<a href="https://agentmods.dev/skills/openlinksoftware/ai-agent-skills/ucp-client"><img src="https://agentmods.dev/badge/skills/openlinksoftware/ai-agent-skills/ucp-client/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/openlinksoftware/ai-agent-skills/ucp-client"><img src="https://agentmods.dev/badge/skills/openlinksoftware/ai-agent-skills/ucp-client.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector warn
SkillSpector: 5 findings, up to high
These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →
- high Privilege Escalation · line 32 Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.Fix: Remove references to credential paths. Use environment variables or secrets managers. For docs, use placeholder paths (e.g., /path/to/config). Never load .env or token files in production code paths.
- high Privilege Escalation · line 144 Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.Fix: Remove references to credential paths. Use environment variables or secrets managers. For docs, use placeholder paths (e.g., /path/to/config). Never load .env or token files in production code paths.
- medium Agent Snooping · line 32 Skill enumerates or reads other installed skills. Access to other skills' SKILL.md files or the skills directory reveals prompt instructions, capabilities, and secrets that should be invisible to peer skills.Fix: Remove all code or instructions that list or read other skills' files or directories. Skills should operate independently; cross-skill access is a privilege escalation.
- medium Agent Snooping · line 126 Skill enumerates or reads other installed skills. Access to other skills' SKILL.md files or the skills directory reveals prompt instructions, capabilities, and secrets that should be invisible to peer skills.Fix: Remove all code or instructions that list or read other skills' files or directories. Skills should operate independently; cross-skill access is a privilege escalation.
- medium MCP Rug Pull · line 96 npx commands without a version suffix (e.g. @1.0.0) create a rug-pull risk if the upstream server is compromised and publishes a malicious update.Fix: Pin the version: npx @scope/[email protected]
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 | $0.00108 | $0.03980 |
| Opus 5 | $0.00054 | $0.01990 |
| Sonnet 5 | $0.00022 | $0.00796 |
| Haiku 4.5 | $0.00011 | $0.00398 |
Grade A, and why
ucp-client scanned grade A with 1 finding 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 6d 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.
Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
1. Accept a protected `resource_url` and establish the request identity without putting secrets on the command line. `scripts/ucp_resource_client.py` accepts a WebID-TLS/NetID client certificate directly, either as PKCS# How it starts
The opening of the file, as written. The whole thing — 145 lines — stays where its author put it; the contents beside it link to each section on GitHub.
UCP Client
Use the resource URL as the primary identifier. Keep RDF as the authoritative offer description, UCP as the commerce/checkout protocol, MPP as the HTTP 402 payment protocol, and Stripe as a payment processor. Do not collapse these layers.
Two entry points
Everything below (steps 1-9) is the --resource-url-driven path: start from a protected resource, work backward through identity/402/RDF discovery to find its offer. When the merchant instead publishes a product feed (RSS 2.0 + Google Merchant g: namespace, e.g. https://<shop>/shop/feed?rss), skip all of that: --product-feed URL fetches the feed directly — it is already a complete, typically-public, pre-authenticated list of every purchasable offer with its own IRI (link/guid), price, and currency. With no --feed-item-id/--feed-search selector the client lists every item and stops so a human or calling agent can pick one; with a selector matching exactly one item, it goes straight to UCP discovery and checkout (steps 6-9 below), skipping the identity-probe/401/402/RDF/SPARQL steps (1-5) entirely — there is no resource ACL to probe and no offer to discover, the feed already answered both.
# List every offer in the shop's feed
python scripts/ucp_resource_client.py --product-feed "https://ods-qa.openlinksw.com/shop/feed?rss"
# Pick one by title search, preview the checkout without creating it
python scripts/ucp_resource_client.py \
--product-feed "https://ods-qa.openlinksw.com/shop/feed?rss" \
--feed-search "Hotpot" --bearer-token-env UCP_BEARER --dry-run
# Same, but actually create + complete the checkout via a Stripe test SPT
python scripts/ucp_resource_client.py \
--product-feed "https://ods-qa.openlinksw.com/shop/feed?rss" \
--feed-item-id "ODSQA-FA-PROPERLANCASHIREHOTPOTRECIPE-0001" \
--bearer-token-env UCP_BEARER --complete-with-stripe-spt
Core workflow (--resource-url path)
- Accept a protected
resource_urland establish the request identity without putting secrets on the command line.scripts/ucp_resource_client.pyaccepts a WebID-TLS/NetID client certificate directly, either as PKCS#12 (--client-p12 PATH --client-p12-password-env ENV_VAR— the bundle is decrypted withcryptographyand staged as temporary0600PEM files, cleaned up on exit; confirmed live against ODS-QA) or PEM (--client-cert/--client-key); for the selected ODS-QA direct-execution profile withcurlitself rather than this script, themtls-curlskill is the alternative. The client also accepts an OAuth access token (Bearer or DPoP), Digest credentials, environment-backed headers, or an already-established ambient identity session. Discover OAuth metadata from the resource origin's OAuth 2.0 Authorization Server Metadata endpoint when using OAuth; the server OAuth backend is assumed to already exist and is not implemented by this skill. - Probe the resource before commerce. Interpret the server response as an identity-first state machine:
401means authentication is missing or failed and may advertise one or more authentication schemes;200means the authenticated identity passes the resource ACL;402means the authenticated identity failed the ACL and payment can grant access;403is a terminal policy denial without a payable challenge. A401may therefore trigger Digest, WebID-TLS/mTLS, or OAuth token acquisition; it is never itself a payment challenge. - Accept
402only when identity was established and the response includes at least oneWWW-Authenticate: Paymentchallenge. Parse repeatedWWW-AuthenticateandLinkfields, preserving Payment challenge parameters,Payment-Receipt,Location, content type, and links such asrel=offer,describedby,alternate, orpayment. A naked402, or402before identity, is a protocol error. - Query the merchant RDF knowledge graph for a
schema:Offerassociated with the resource IRI. Accept both Schema.org namespace forms (http://schema.org/andhttps://schema.org/). Match directschema:itemOffered,schema:url,schema:contentUrl, orschema:identifierlinks and item-mediated links such as Offer → License/Product → resource. Use--resource-predicate IRIfor additional explicit merchant relations. Use--sparql-endpointwhen known; otherwise try<merchant-origin>/sparql; then use response-linked or explicit RDF metadata. Quad stores commonly keep offer data in a named graph outside the SPARQL protocol default graph: an unscoped query that returns no rows is automatically retried once, scanned across all named graphs (GRAPH ?g { ... }), before falling back to RDF dereference; pass--sparql-default-graph IRI(repeatable) to scope explicitly instead of relying on the automatic scan. If the resource's public/RDF-published IRI differs from the URL actually used to access it (see the ODS-QA case below), pass--match-urlwith the RDF-published IRI so offer matching succeeds independently of the access URL. If discovery still finds no offer at all (some MPP/x402-gated resources publish no RDF description whatsoever), fall back to decoding the offer identity directly out of the resource's ownWWW-Authenticate: Payment request=<base64url JSON>challenge parameter, when present, rather than failing outright. - Extract offer identity, item identity, price, currency, availability, seller, and UCP item identifier. Read price/currency either directly or through
schema:priceSpecification. Preferschema:sku, thenschema:productID, then literalschema:identifier; known explicit merchant identifiers such asoplofr:offerNumbermay follow. For another merchant vocabulary use repeatable--item-id-predicate IRI, or supply a verified--item-id VALUE. Never infer an ID from a URL path.schema:potentialActionquery parameters remain opt-in via--allow-action-item-id. - Discover the merchant UCP profile at
/.well-known/ucpand read the advertiseddev.ucp.shoppingREST checkout capability/version. - Create a UCP checkout session using the mapped item ID and quantity. Treat the checkout response as authoritative for final price, totals, eligibility, and status. Before spending money, surface the checkout amount/currency and payment boundary unless the user explicitly authorized the purchase and any required spending limit is satisfied.
- Pay for the checkout through whichever surface the merchant actually settles it on — these are two distinct targets, not interchangeable:
- UCP checkout completion: when the merchant advertises a payment handler compatible with a Shared Payment Token (e.g.
opl_shop_stripe_spt, following the Link Agent Wallet shape from Stripe's UCP payments handler), pass--complete-with-stripe-sptto fetch a Stripe test-mode SPT capped at the checkout's authoritative total andPOST .../completedirectly — this finishes the UCP checkout itself and returns the resulting order. - Resource-level MPP payment: fulfill one server-advertised Payment challenge on the protected resource using an MPP implementation such as
mppx(--mpp-command). Preserve identity independently through WebID-TLS/mTLS, an authenticated session cookie, or a server-side challenge binding; Basic/Digest/Bearer/DPoP and Payment cannot both occupy the singleAuthorizationfield. After entitlement binding, retry the resource with the original identity context and capturePayment-Receiptplus the final status.
- UCP checkout completion: when the merchant advertises a payment handler compatible with a Shared Payment Token (e.g.
- Reconcile an MPP receipt with the UCP checkout only through a merchant-supported binding — the two are not automatically the same transaction. Return the resource plus provenance: identity-gate state, resource URL, offer IRI, checkout ID/status, amount/currency, payment method/intent, receipt or order, and final HTTP status.
What ships with it
5 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
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.
- 6d ago First seen · 145 lines · 108 tokens per session scan A e09587d971c8
ucp-client is a skill published in the GitHub repository OpenLinkSoftware/ai-agent-skills (38 stars, last pushed yesterday), licensed MIT. It adds 108 tokens to every session and 3,980 once invoked, about $0.0005 per session on Opus 5. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-09-04.
Other skills, from other repositories
amazon-reviews-api-skill
This skill helps users automatically extract Amazon product reviews via the Amazon Reviews API. Agent should proactively apply this skill when users express needs like getting reviews for Amazon product with ASIN B07TS6R1SF, analyzing customer feedback for a specific Amazon item, getting ratings and comments for a…
amazon-competitor-analyzer
Scrapes Amazon product data from ASINs using browseract.com automation API and performs surgical competitive analysis. Compares specifications, pricing, review quality, and visual strategies to identify competitor moats and vulnerabilities.
asc-subscription-localization
Bulk-localize subscription, subscription-group, and in-app purchase display names across App Store locales using asc, including API 4.4.1 version-scoped v2 resources. Use when filling or updating subscription/IAP names and descriptions without App Store Connect UI work.
food-order
Reorder previous Foodora orders, preview cart contents, and track delivery ETA/status with ordercli. Use when the user wants to reorder food, check delivery status, or browse recent Foodora order history. Never confirm an order without explicit user approval.
product-description-generator
E-commerce product description generator for any platform. Generates optimized titles, bullet points, descriptions, and backend keywords using competitor research + keyword scoring + FABE copywriting. Two modes: (A) Create — generate listing from product specs with optional competitor analysis, (B) Optimize — improve…
amazon-price-tracker
Amazon price monitoring and competitive pricing intelligence. Real-time price tracking, Buy Box analysis, promotion detection, and dynamic pricing strategy optimization. Use when the user asks about price monitoring, competitor pricing, Buy Box tracking, or pricing strategy.