ByHeartAI
Advanced8 min read

Design a Coding Agent

A coding agent reads and patches a repo through tools, runs tests in a jail, and ships a PR for a human. Dumping the whole codebase into context is not a design; it's a bill and a leak.

Explain like I'm new to AI

Humans don't memorize the company monorepo to fix a bug. Neither should the model. Explore → edit → verify → review.

Tests (or typecheck) in a sandbox. The verifier is code, not another model's vibes. Loop cap.

A coding agent is tools plus a sandbox plus tests — the PR is HITL. The repo is not a 2-million-token paste.

Job: produce a change that passes checks for this task. SLO: tests green, limited steps, no secret exfil, p95 that a developer will wait (or overnight batch).

Shape: agent (the path through the repo is unknown) with a hard cap and a verifier.

Mental model

A contractor with a badge to the repo, a laptop that can't see ~/.aws, and a rule: nothing merges without CI + a person.

How it works

  • Tools: list, grep/search, read, apply_patch, run_tests. Not unrestricted shell if you can avoid it.
  • Context: retrieve/grep first (code RAG). Pack only the files that matter (context engineering).
  • Verify: unit tests, typecheck, linters — code as judge (evaluation / GRPO-style when you train; at serve time it's just CI).
  • Sandbox: no home directory, egress allow-list, no Docker socket (secrets-sandboxing).
  • HITL: the pull request. Show the diff. The model cannot click merge.
  • Eval: held-out tasks (issue → failing test → patch). Trajectory eval, not BLEU on the chat.
  • Loop cap and $ cap (unbounded consumption).

Real-world example

Agent "fixes" auth by printing a key it found in .env into the reply. Jail + don't index secrets + output filters. Tests wouldn't catch the leak — security eval would.

Technical explanation

Long-context models still lose the middle of a 400-file dump. Tools beat paste.

MCP for repo tools is fine; treat the server like installable software (MCP security).

Multi-agent "writer vs reviewer" can help if the reviewer is tests + linter, not another ungrounded LLM arguing.

Common mistakes

Common mistake

Pasting the repository into the prompt because the window is "a million tokens."

  • exec on the developer machine.
  • No tests, only "looks like good code."

When to use it

  • Well-tested repos, scoped tasks, humans who will read PRs.

When NOT to use it

  • Don't unleash write tools on prod. Don't skip sandbox "because it's local."
  • Greenfield "build me an app" with no tests is a generator, not this design.

Alternatives

  • Completions in the editor (chatbot). Workflow: generate patch from a fixed template. Human pair-programming.

Quick quiz

Question 1 of 3

How should a coding agent see the repo?

Question 2 of 3

What verifies the change?

Question 3 of 3

True or false: a million-token window means you should dump the whole repository every turn.

Related concepts

  • Secrets Management and SandboxingKeep keys out of prompts, run untrusted work in a jail with an egress allow-list, and cap loops and spend so agents cannot become a denial-of-wallet.
  • Human-in-the-LoopHuman-in-the-loop inserts a person's approval before risky or irreversible agent actions, combining autonomy with safety and accountability.
  • Planning & ReflectionPlanning breaks a goal into sub-tasks before acting, and reflection lets the agent critique and fix its own work — together they make agents far more reliable.
NextDesign a Research or Multi-Agent System

Last reviewed: 2026-09-04 · Written by ByHeart AI · Reviewed by ByHeart AI