Getting it into your agent
It runs from inside its repository, so the clone comes first — what it calls does not travel with the file alone.
git clone --depth 1 https://github.com/epistemedeus/x402-url-extractornpx agentmods add skills/epistemedeus/x402-url-extractor/consumerWrote 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/epistemedeus/x402-url-extractor/consumer)<a href="https://agentmods.dev/skills/epistemedeus/x402-url-extractor/consumer"><img src="https://agentmods.dev/badge/skills/epistemedeus/x402-url-extractor/consumer/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/epistemedeus/x402-url-extractor/consumer"><img src="https://agentmods.dev/badge/skills/epistemedeus/x402-url-extractor/consumer.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 | $0.00030 | $0.00868 |
| Opus 5.5 | $0.00012 | $0.00347 |
| Sonnet 5.5 | $0.00006 | $0.00174 |
| Haiku 4.5 | $0.00003 | $0.00087 |
Grade A, and why
seller-repair-external-consumer 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 4d 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.
What it actually says
Seller repair external consumer
Use this when a visiting agent brings their own authorized target, operation, expected output, and SDK. package/cases/retained-case.json replays one loopback catalog fixture. It is not the only task the caller may submit.
The executable is package/bin/seller-repair.mjs version 0.2.0. A later process uses package/bin/cold-later.mjs. Pins for the vendored privacy files are package/references/pins.json. The license is package/LICENSE. The acquisition note is package/SOURCE-NOTICE.txt.
Runtime dependencies are Node.js >=22.22.0 and node:crypto, node:dns/promises, node:fs, node:fs/promises, node:http, node:https, and node:net. There is no npm dependency. Unpack with system tar and gzip. Do not add node_modules, .git, or .grok.
From the extract root:
node package/bin/seller-repair.mjs deliver
node package/bin/seller-repair.mjs diagnose --case package/cases/supplied-quota.json
node package/bin/seller-repair.mjs diagnose --case package/cases/supplied-health.json
node package/bin/seller-repair.mjs reproduce --case package/cases/retained-case.json
node package/bin/seller-repair.mjs retest --case package/cases/retained-case.json --mode repaired
node package/bin/cold-later.mjs package/cases/retained-case.json --resource /catalog/items
node package/bin/seller-repair.mjs reject-seeded package/fixtures/seeded-false-useful.json
node package/bin/seller-repair.mjs readonly --case package/cases/public-readonly.json
node package/bin/seller-repair.mjs receive --case package/cases/retained-case.json --base http://127.0.0.1:PORT --retest-base http://127.0.0.1:PORT
deliver exits 0 when the offline health evidence is useful, the bare 200 is not, the private address is refused, and the loopback direct probe agrees. diagnose reads supplied evidence and does not purchase. reproduce exits 0 when the loopback retest matches. That result is a loopback fix. deployedCounterpartyRepair stays false. reject-seeded and a stale later binding exit 2. Exit 0 is not a payment and is not proof that an unknown report was useful.
Only GET is probed. probeConsent.confirmed must be true. public-https names one public HTTPS origin. public-read-only may name only https://agents.samedaydesk.com and /.well-known/public-acquisition/index.json. loopback may name only http://127.0.0.1 with a port. Private and reserved addresses and a changed DNS answer are refused before a connection. Redirects stay 0. A trickling body cannot renew deadlineMs. HTTP 200, a declaration, unknown coverage, and a missing field do not prove useful execution or a paid need. The existing $0.01 seller-integrity audit is not purchased. An incomplete local audit remains a useful negative. Tokens and adaptation cost stay unknown.
This skill is unlaunched. Root hosts the 0.2.0 archive and runs candidate/cold-command.json on those hosted bytes. The 0.1.0 archive bytes stay frozen.
What ships with it
6 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.
- 4d ago First seen · 40 lines · 30 tokens per session scan A 765d29d4b491
seller-repair-external-consumer is a skill published in the GitHub repository epistemedeus/x402-url-extractor (1 stars, last pushed today), licensed MIT. It adds 30 tokens to every session and 868 once invoked, about $0.0001 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
x402
Monetize any user project/service with the x402 payment protocol on platform networks (Base + Monad + Robinhood + X Layer + Solana; Starchild platform billing: payperuse / lifetime / weekly / monthly / quarterly / yearly / prepaid, plus multi-plan services), limited-time free promotions (amount-0 verify, no…
config-loader-helper
Diagnose configuration-related failures, enumerate required env vars, and guide safe local test setup (no secrets).
fix-suggester
Diagnose failures and propose minimal, test-backed fixes with verification and rollback instructions.
agoragentic-transaction-assurance-eval
Evaluate autonomous agents against deterministic Transaction Assurance and Tumbler marketplace scenarios covering mandates, provider choice, budgets, payments, retries, delivery, outcomes, refunds, and reconciliation. Use to produce public-safe evidence before enabling economic autonomy.
agent-payments-assurance-challenge
Validate and score a bounded, self-attested offline payment-assurance run record against the unpublished alpha challenge contract. Use for local conformance regression evidence, not certification or production-readiness claims.
bankr-shopify
Shopify Admin & Storefront GraphQL APIs via curl, with Bankr-native bridges. Manage products, orders, customers, inventory, metafields, webhooks, and bulk ops, then wire merchant data to onchain primitives — store a Bankr-resolvable handle (ENS, Twitter, Farcaster, wallet) on each customer as a metafield, expose…