Agent-building steps 3 & 4: turn fixed procedures into tools for determinism; .md files often enough for memory
PawelHuryn · x · 2026-09-09
Paweł Huryn's agent-building series continues. Step 3 (connect tools): inventory everything the agent needs — simple tools, custom functions, integrations, CLIs, data sources, MCP servers. Key tip: turn every fixed-step procedure into a tool to make it deterministic.
Step 4 (memory): agents must track progress and learn; options include short-term memory (variables, recent interactions), long-term memory (files, vector, SQL, graph), and a scratchpad ('think tool'). In many cases an agent using .md files suffices.
Related event: Practical Guide to Building AI Agents: Orchestration, Memory and Tools(3 posts)→
More from coding & agent
- Kolab Benchmark Uses Pro Screen Recordings to Show AI Agents Overestimate Collaboration Quality — minnesotanlp · 2026-10-27
- Dev argues Claude burns far more tokens than GPT models for the same coding work — haider1 · 2026-10-27
- Stop building agents that can't change models: stable contracts, external state, real evals — rseroter · 2026-10-27
- Cheap models can cost more in agent harnesses than frontier ones, researcher warns — vasuman · 2026-10-27
- Kanzo: A 5k-Line Rust Kanban Binary With MCP for Claude Code and Codex — allmudi · 2026-10-27
- 'A human approves everything' is a trap: approval fatigue and how to actually design safe agents — Efistoffeles · 2026-10-27