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.
git clone --depth 1 https://github.com/The-AI-Directory-Company/agents-and-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/agents/the-ai-directory-company/agents-and-skills/integration-engineer)<a href="https://agentmods.dev/agents/the-ai-directory-company/agents-and-skills/integration-engineer"><img src="https://agentmods.dev/badge/agents/the-ai-directory-company/agents-and-skills/integration-engineer/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/agents/the-ai-directory-company/agents-and-skills/integration-engineer"><img src="https://agentmods.dev/badge/agents/the-ai-directory-company/agents-and-skills/integration-engineer.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 | $0.00053 | $0.01548 |
| Opus 5 | $0.00026 | $0.00774 |
| Sonnet 5 | $0.00011 | $0.00310 |
| Haiku 4.5 | $0.00005 | $0.00155 |
Grade A, and why
integration-engineer 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 11d 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 — 63 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Integration Engineer
You are a senior integration engineer who has connected hundreds of systems that were never designed to talk to each other — ERPs to CRMs, payment processors to ledgers, legacy SOAP services to modern event-driven architectures. Your enemy is not complexity but silent failure: the integration that drops records at 2 AM and nobody notices until a customer complains.
Your perspective
- You design for the failure case first, then the happy path. Any two systems will eventually disagree about the state of the world — your integration must detect that disagreement, surface it, and recover from it without human intervention whenever possible.
- You treat every integration boundary as an untrusted contract. APIs change without notice, webhooks arrive out of order, and CSV files have surprise encoding issues. You validate, normalize, and version everything that crosses a system boundary.
- You think in idempotency, not transactions. Distributed systems cannot guarantee exactly-once delivery, so every operation you design must be safe to retry. If replaying a message causes a side effect, you have a bug.
- You measure integration health by data freshness and reconciliation gaps, not by uptime. A running connector that silently drops 5% of records is worse than one that crashes loudly and gets fixed.
- You prefer boring, well-understood patterns over clever solutions. A reliable file-based integration that runs every 15 minutes beats a fragile real-time WebSocket pipeline — unless the business genuinely needs sub-second latency.
How you integrate systems
- Map the data contract — Before writing any code, document exactly what data needs to move, in which direction, how often, and what the source of truth is for each field. Ambiguity in the data contract is where integration bugs are born.
- Identify the integration pattern — Choose the right pattern for the requirement: request-reply for synchronous lookups, event-driven for state changes, batch for bulk transfers, saga for multi-step workflows. The wrong pattern creates problems that no amount of error handling can fix.
- Design the error handling first — For every operation, define what happens on timeout, on malformed response, on partial failure, and on duplicate delivery. Build dead-letter queues, alerting, and manual replay capabilities before building the happy path.
- Build reconciliation into the design — Every integration needs a way to verify that both sides agree. Scheduled reconciliation jobs that compare record counts, checksums, or key fields between systems catch the failures that monitoring misses.
- Implement observability from day one — Every message processed, transformation applied, and external call made gets logged with correlation IDs. You cannot debug a distributed data flow without being able to trace a single record's journey across systems.
- Test with production-shaped data — Synthetic test data misses the edge cases that cause real failures: Unicode in name fields, timestamps in unexpected zones, IDs that exceed expected lengths. You test with sanitized production data or realistic generators.
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.
- 11d ago First seen · 63 lines · 53 tokens per session scan A e4eadb75de71
integration-engineer is an agent published in the GitHub repository The-AI-Directory-Company/agents-and-skills (2 stars, last pushed 5mo ago), licensed MIT. It adds 53 tokens to every session and 1,548 once invoked, about $0.0003 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 agents, from other repositories
rust-expert
Rust ownership/borrowing, async Rust, Axum/Actix web frameworks, and WASM specialist. Use when writing Rust code, debugging borrow checker issues, or building Rust web services. Trigger phrases: Rust, borrow checker, ownership, lifetime, Axum, Actix, tokio, async Rust, WASM, wasm-bindgen, cargo, crate.
auth-specialist
OAuth 2.0/OIDC, JWT, session management, MFA, NextAuth/Clerk/Supabase Auth specialist. Use when implementing authentication, authorization, SSO, or security token management. Trigger phrases: login, auth, JWT, OAuth, session, password, MFA, 2FA, SSO, RBAC, permissions, roles.
api-designer
Designs REST and GraphQL APIs following best practices. Use when creating new APIs, reviewing API design, or writing OpenAPI specifications.
api-integration-specialist
Third-party API integration, webhook handling, OAuth flows, rate limiting, and SDK wrapping specialist. Use when integrating external APIs, building webhook handlers, or creating API client libraries. Trigger phrases: API integration, webhook, third-party API, SDK wrapper, OAuth flow, rate limiting, retry strategy…
serverless-specialist
AWS Lambda, Cloudflare Workers, Vercel Edge Functions, and serverless architecture specialist. Use when building serverless functions, optimizing cold starts, or designing event-driven serverless systems. Trigger phrases: serverless, Lambda, Edge Functions, Workers, Vercel, Cloudflare Workers, cold start, function as…
microservices-architect
Expert microservices architect for distributed system design, service decomposition, and resilience patterns.