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 instructions/paymobaccept/paymob-ai-integration-skill/agents-mdgit clone --depth 1 https://github.com/PaymobAccept/Paymob-AI-Integration-SkillWrote 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/instructions/paymobaccept/paymob-ai-integration-skill/agents-md)<a href="https://agentmods.dev/instructions/paymobaccept/paymob-ai-integration-skill/agents-md"><img src="https://agentmods.dev/badge/instructions/paymobaccept/paymob-ai-integration-skill/agents-md.svg" alt="Measured on agentmods" 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 | $0.01152 | $0.01152 |
| Opus 5 | $0.00576 | $0.00576 |
| Sonnet 5 | $0.00230 | $0.00230 |
| Haiku 4.5 | $0.00115 | $0.00115 |
Grade A, and why
Paymob-AI-Integration-Skill AGENTS.md 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 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.
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 — 48 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AGENTS.md — Paymob Payment Integration
This repository is a portable skill that makes any AI coding agent an expert at integrating the Paymob payment gateway (Egypt, UAE, KSA, Oman). It is not an app to build — it is guidance to apply to the user's project.
Only have this file? The full skill (workflow + per-stack code + references) lives at https://github.com/PaymobAccept/Paymob-AI-Integration-Skill. When a path referenced below is not present in the current project, fetch it from that repo — raw base:
https://raw.githubusercontent.com/PaymobAccept/Paymob-AI-Integration-Skill/main/.
AI agents: use the following authority order:
- For volatile endpoint shapes, HMAC fields, SDK versions, and regional behavior, the current official Paymob docs referenced by
skills/paymob-integration/references/live-resources.mdwin. - For the packaged workflow,
skills/paymob-integration/SKILL.mdand its task-specific references are canonical. universal-prompt.mdis the self-contained fallback for hosts that cannot load the skill package; keep it synchronized with the canonical skill.- For live account actions, Paymob's official MCP server is
https://mcp.paymob.com/mcp. Setup and its versioned tool catalog are inskills/paymob-integration/references/mcp-server.md. It never replaces the HMAC-verified webhook as the source of truth.
Multi-agent work
For broad integrations or audits, delegate independent read-only tasks such as codebase mapping, current-doc verification, and security review. Subagents receive no Paymob credentials, do not use authenticated Paymob tools, and return findings only with file/line references. Keep one primary agent responsible for requirements, final edits, tests, and every live action. If edits must be delegated, give agents exclusive non-overlapping paths and merge through the primary agent.
Non-negotiable rules
- Intention API only. Create payments with
POST {base_url}/v1/intention/. Never use or suggest the legacy 3-step flow (auth token → order → payment key). - HMAC is always SHA-512. Concatenate the documented fields in the exact order, hex-lowercase, and compare timing-safely. There are 3 HMAC types (transaction = 20 fields, card token = 8 fields, subscription = string formula) — use the right one per callback.
- The HMAC-verified webhook callback is the source of truth for payment status — never the browser
redirection_urlparams (unauthenticated) or a mobile SDK result (UX only). Deduplicate callbacks on unique Paymob transaction/event ID (obj.id); in one database transaction, compare-and-set the order state and insert a uniquely keyed fulfillment outbox record. Useorder.id/special_referenceonly for correlation. - Amount is in the smallest currency unit (cents/piasters): 100.00 EGP =
10000. - Post-payment auth uses the header
Authorization: Token {secret_key}— neverBearer, neverauth_tokenin the body. Transaction Inquiry is the documented API-Key → short-livedAUTH_TOKENexception. - Secrets stay server-side. Only the Public Key (
pk_*) is safe in frontend code. Never expose or commit the Secret Key, API Key, or HMAC Secret; never create the intention from a browser or mobile app. - Prefer a prebuilt integration when one exists. Shopify → a Paymob app; WooCommerce/Magento/Odoo/OpenCart/PrestaShop/… → Paymob's official plugin. Only hand-code for custom/headless checkouts.
- Every live write needs current, specific authorization. Obtain explicit confirmation for the current account, test/live mode, operation, target, amount, and currency; never reuse blanket approval. Read remote state first and keep a stable operation fingerprint/merchant reference.
- Never auto-retry an ambiguous financial write. After a timeout or unclear response, query Paymob to determine whether it succeeded before retrying. Verify and report the remote result after every write.
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 · 48 lines · 1,152 tokens per session scan A 1e81d951b298
Paymob-AI-Integration-Skill AGENTS.md is an instructions file published in the GitHub repository PaymobAccept/Paymob-AI-Integration-Skill (30 stars, last pushed 21d ago), licensed MIT. It adds 1,152 tokens to every session, about $0.0058 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 instructions, from other repositories
wealth-forge-ai CLAUDE.md
Claude Code instructions for vikisingh23/wealth-forge-ai, covering wealthforge ai: claude instructions, persona, user profile, global mandates and agent specialization mapping.
wealth-forge-ai AGENTS.md
AGENTS.md instructions for vikisingh23/wealth-forge-ai, covering wealthforge ai agents, core utilities, the workforce (team wealthforge), 🧠 the brain and 👤 the profile.
wealth-forge-ai GEMINI.md
Gemini CLI instructions for vikisingh23/wealth-forge-ai, covering wealthforge ai, agent instructions, financial rules & standards and configuration.
claude-investment-skills CLAUDE.md
Claude Code instructions for ssurmic/claude-investment-skills: Claude Code reads this file by convention. The canonical guide is AGENTS.md — same content, broader convention name (Codex / Cursor / generic agents all look for AGENTS.md).
Vibe-Research AGENTS.md
AGENTS.md instructions for simonlin1212/Vibe-Research, covering vibe-research-agent 金融研究宪法, 0. 三条不可越线, 1. 数据纪律(五问 gate:给出任何数字或结论前逐条自问), 2. 研究哲学(评估框架,不是预测工具) and 3. 估值口径(你只选输入、解释输出;计算交给 calc/).
daily-watchlist CLAUDE.md
Instructions for Benboerba620/daily-watchlist: This file is the source protocol for the Daily Watchlist workflow. Hypothesis Tracker is bundled in this repository as the thesis/evidence layer.