We built an internal assistant that researches crypto markets and drafts a briefing for analysts, using a planner and specialised sub-agents. The lessons: isolating sub-agent context, a file scratchpad instead of one huge conversation, cost and tool caps, and evals. It does research, never trading.
Agent skills, plugins and MCP servers extend coding and productivity agents with instructions and scripts. Installing one gives its author influence over what your agent reads, writes and runs. The risks we found reviewing community skills, a review checklist, and the sandboxing and policy controls we use.
Users of our internal assistant kept repeating the same preferences: their team, report format, time zone. Adding memory sounds simple and is not. How we separated short-term and long-term memory, decided what to store, handled conflicts and stale facts, and gave users control over what the assistant remembers.
Connecting WhatsApp, Telegram or Slack to an LLM agent that can run tools on your server is a few hundred lines of glue and a large attack surface. This is the architecture we use: normalised channel adapters, sender pairing, a locked-down tool sandbox and hard budget caps.
Most agent demos are one big loop around a model. In production we replaced ours with four explicit patterns: a router, a planner with a fixed step budget, a bounded reflection pass and an approval gate for side effects. Here is how each works and where it fails.
AI coding agents kept running the wrong test command, editing generated files and reading our .env. The fix was repo setup, not better prompts: a short instruction file, deterministic commands, permission rules, formatting hooks and a checked-in MCP config.
Teams wanted AI assistants to answer questions about internal ticket and order data through the Model Context Protocol. A local stdio server is easy; a shared remote one that respects each user's permissions is not. Our design for authentication, authorization, read-only tools, limits and audits.
A step-by-step build of a small Model Context Protocol server in Python with tools, resources and prompts, then how we test it three ways: the MCP Inspector, a scripted client in pytest and a real AI client, before anyone on the team relies on it.