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 jpmorgan-payments/pdp-skills --skill jpm-oauthgit clone --depth 1 https://github.com/jpmorgan-payments/pdp-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/jpmorgan-payments/pdp-skills/jpm-oauth)<a href="https://agentmods.dev/skills/jpmorgan-payments/pdp-skills/jpm-oauth"><img src="https://agentmods.dev/badge/skills/jpmorgan-payments/pdp-skills/jpm-oauth/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/jpmorgan-payments/pdp-skills/jpm-oauth"><img src="https://agentmods.dev/badge/skills/jpmorgan-payments/pdp-skills/jpm-oauth.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 | $0.00170 | $0.02496 |
| Opus 5 | $0.00085 | $0.01248 |
| Sonnet 5 | $0.00034 | $0.00499 |
| Haiku 4.5 | $0.00017 | $0.00250 |
Grade A, and why
jpm-oauth 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 8d 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 — 153 lines — stays where its author put it; the contents beside it link to each section on GitHub.
JPM OAuth implementation
Generate JWT-signing + access-token-exchange code that follows JPM's IDAnywhere OAuth contract. The generated module caches the access token until just before its server-set expiry, so the merchant's app reuses one token across many API calls instead of regenerating per request.
This skill is typically invoked from the end of jpm-integrations-get-started (its Step 4). When run standalone, it gathers the same inputs interactively.
Step 1 — Verify prerequisites
The merchant must have all four of:
- clientId (e.g.
CC-123456-D12992-232222-TEST) - certificate path (PEM
.cerfrom JPM) - private key path (PEM key paired with the certificate)
- SHA-1 thumbprint of the cert (40 hex chars; colons optional — code will normalize)
If you arrived from jpm-integrations-get-started, these are already in conversation context — use them.
If invoked standalone:
- Ask the user where the credentials live: file paths, an
.env, or in conversation. - If anything is missing, tell the user to run
jpm-integrations-get-startedfirst and exit. Do not invent placeholder values.
Step 2 — Environment
Ask:
- Question: "Which JPM environment are you targeting?"
- Header: "Environment"
- Options:
- "CAT (test)"
- "PROD"
- "Both — generate code that switches via env var"
Do not prompt the user for the resource_id. The generated module reads it from the JPM_RESOURCE_ID environment variable at runtime, and the intake skill's .env.example already includes that key with a comment pointing the user at the JPM onboarding email (where the value lives). Just remind the user in your wrap-up to fill JPM_RESOURCE_ID in their .env before running the auth module — and that CAT and PROD have different resource_id values, so the right one must be in the .env for the environment they're targeting.
Step 3 — Education and JWT TTL
Show the user this guidance verbatim before asking the TTL question:
Two TTLs to know about:
Access token TTL — set by JPM's IDAnywhere server, returned as
expires_in(seconds) in every token response. The generated module reads this value and reuses the access token across requests until shortly before it expires. Don't treat it as fixed — server-side defaults can change.JWT TTL — client-controlled (you pick). The JWT is the credential you hand to IDAnywhere to get an access token. JPM's guidance: 8 hours for PROD; longer is fine for CAT.
Anti-pattern to avoid: generating a fresh JWT and a fresh access token on every API call (think of it like running a card auth on a brand-new payment instrument every transaction — wasteful and slow). The access token is reusable until expiry. The generated module caches it in memory and only refreshes it when within ~30s of expiry. Call
getAccessToken()once per outbound request and let the cache do its job.
What ships with it
4 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.
- 8d ago Changed · +2 lines 77579312bdb9
- 12d ago First seen · 151 lines · 170 tokens per session scan A 5c9d76359fa2
jpm-oauth is a skill published in the GitHub repository jpmorgan-payments/pdp-skills (8 stars, last pushed yesterday), licensed Apache-2.0. It adds 170 tokens to every session and 2,496 once invoked, about $0.0009 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 2026-08-31.
Other skills, from other repositories
messari-x402
Access Messari's full API via x402 pay-per-request — no API key needed. Asset data, market metrics, signals, news, fundraising, token unlocks, on-chain networks, and AI chat, all paid with USDC on Base.
sumup-mcp
Use the SumUp MCP server (https://mcp.sumup.com/mcp) from Cursor, Claude Code, Codex, or any MCP-capable client. Use when the user mentions SumUp MCP, needs to wire mcp.sumup.com, or wants tool-based access to SumUp APIs.
payuni-query
A tool for checking PAYUNi transaction records through its Query API. PAYUNi is a Taiwan payment gateway, and a transaction query asks the gateway for an order’s current payment result.
kryptogo-pay-query
An implementation guide for querying KryptoGO Payment, a stablecoin payment service, to check individual or multiple payment records.
x402
Set up Browser Use Cloud payments with x402 — pay per request from a crypto wallet (USDC on Base mainnet), no signup or API key. Two setups it works out up front — "just use it" (set up a wallet so you or Claude Code can run cloud browser tasks paid from the wallet — Claude writes and runs throwaway scripts, nothing…
tushare
A Python interface for Tushare, a financial data service that provides market and company information for stocks, funds, futures, and digital assets. It returns queried data as pandas tables.