--- title: DIY harness slug: diy-harness url: "https://toolweight.com/tools/diy-harness" homepage: "https://toolweight.com/compare/coding-agents" categories: coding-agents last_verified: 2026-11-14 dataset_version: 0738be72 license: CC-BY-4.0 --- # DIY harness > Your own loop over a model API with an agent SDK, the build-it-yourself reference point. - The baseline: a few hundred lines over whichever agent SDK your provider ships (every vendor here has one). - Carries only the tools your codebase needs; this row endorses no specific SDK. - The right answer for narrow, repeated tasks (a migration bot, a test-backfill job). - Wrong for open-ended engineering: you will spend a quarter rebuilding permissions, checkpointing and context management badly. ## Identity | | | | --- | --- | | Name | DIY harness | | Company | DIY harness | | One-liner | Your own loop over a model API with an agent SDK, the build-it-yourself reference point. | | Site | https://toolweight.com/compare/coding-agents | | Open source | Yes | | Compared in | 1 | ## Where it is compared ### [Coding agents](https://toolweight.com/compare/coding-agents) Ranked **#12 of 17** on default weights. | Field | Value | Confidence | Verified | Source | Note | Permalink | | --- | --- | --- | --- | --- | --- | --- | | Surface | CLI | Inferred | - | - | Whatever you build it as. | https://toolweight.com/compare/coding-agents#diy-harness--surface | | Default model | Whichever you wire up | Inferred | - | - | - | https://toolweight.com/compare/coding-agents#diy-harness--default_model | | Model breadth | any provider / router | Inferred | - | - | - | https://toolweight.com/compare/coding-agents#diy-harness--model_breadth | | Repo indexing | manual context only | Inferred | - | - | Until you build something better, which is most of the work. | https://toolweight.com/compare/coding-agents#diy-harness--repo_indexing | | Plan mode | ○ | Inferred | - | - | - | https://toolweight.com/compare/coding-agents#diy-harness--plan_mode | | Checkpoint & undo | ○ | Inferred | - | - | - | https://toolweight.com/compare/coding-agents#diy-harness--checkpointing | | Headless / CI | ● | Inferred | - | - | It is a script; headless is the default. | https://toolweight.com/compare/coding-agents#diy-harness--headless_ci | | Local models | ● | Inferred | - | - | - | https://toolweight.com/compare/coding-agents#diy-harness--local_models | | Where it stops | - | Unknown | - | - | There is no ceiling, it goes exactly as far as you build, and you own every rung. Not scored, because ranking your own code against shipped products would be meaningless. | https://toolweight.com/compare/coding-agents#diy-harness--stops_at | | MCP | ◐ | Inferred | - | - | Most agent SDKs ship an MCP client. | https://toolweight.com/compare/coding-agents#diy-harness--mcp_support | | Subagents | ◐ | Inferred | - | - | - | https://toolweight.com/compare/coding-agents#diy-harness--subagents | | Hooks | ● | Inferred | - | - | It is your code; every line is a hook. | https://toolweight.com/compare/coding-agents#diy-harness--hooks | | Skills / commands | ○ | Inferred | - | - | - | https://toolweight.com/compare/coding-agents#diy-harness--skills_commands | | Background agents | ◐ | Inferred | - | - | - | https://toolweight.com/compare/coding-agents#diy-harness--background_agents | | Opens PRs | ◐ | Inferred | - | - | - | https://toolweight.com/compare/coding-agents#diy-harness--pr_authoring | | Permissions | applies everything | Inferred | - | - | Unless you build a permission layer, which almost nobody does properly. | https://toolweight.com/compare/coding-agents#diy-harness--permission_model | | BYOK | ● | Inferred | - | - | - | https://toolweight.com/compare/coding-agents#diy-harness--byok | | Licence | - | Unknown | - | - | Yours. | https://toolweight.com/compare/coding-agents#diy-harness--licence | | Pricing model | BYOK tokens only | Inferred | - | - | - | https://toolweight.com/compare/coding-agents#diy-harness--pricing_model | | Entry price | $0/seat/mo | Inferred | - | - | Zero licence cost, non-zero engineering cost, budget a fortnight before it is useful and a quarter before it is safe. | https://toolweight.com/compare/coding-agents#diy-harness--entry_price | | Token markup | ○ | Inferred | - | - | - | https://toolweight.com/compare/coding-agents#diy-harness--gateway_markup | | GitHub stars | - | Unknown | - | - | - | https://toolweight.com/compare/coding-agents#diy-harness--github_stars | | Terminal-Bench | - | Unknown | - | - | - | https://toolweight.com/compare/coding-agents#diy-harness--terminal_bench | | Bench model | - | Unknown | - | - | - | https://toolweight.com/compare/coding-agents#diy-harness--terminal_bench_model | | Positioning | The reference point: what you get for a few hundred lines and an API key. | Inferred | - | - | - | https://toolweight.com/compare/coding-agents#diy-harness--positioning | | First release | - | Unknown | - | - | - | https://toolweight.com/compare/coding-agents#diy-harness--first_release | **Verdict.** Right for narrow, repeated jobs, wrong for open-ended engineering. ## Alternatives - [Aider](https://toolweight.com/tools/aider), Minimal terminal pair-programmer with a tree-sitter repo map and automatic git commits. - [Amp](https://toolweight.com/tools/amp), Sourcegraph's opinionated agent with no model picker and an ad-supported free tier. - [Claude Code](https://toolweight.com/tools/claude-code), Anthropic's terminal-first agent, with IDE, web and GitHub Action surfaces. - [Cline](https://toolweight.com/tools/cline), Open-source VS Code and JetBrains agent built around an explicit plan/act split. - [Codex CLI](https://toolweight.com/tools/codex-cli), OpenAI's open-source Rust agent, with the strongest local sandboxing on this page. - [Cursor](https://toolweight.com/tools/cursor), VS Code fork with a real codebase index, background cloud agents and a CLI. - [Devin](https://toolweight.com/tools/devin), Fully delegated cloud engineer: file a task in Slack or the web app, get a PR back. - [Gemini CLI](https://toolweight.com/tools/gemini-cli), Apache-2.0 terminal agent from Google with an unusually generous free tier. - [GitHub Copilot (Agent HQ)](https://toolweight.com/tools/github-copilot), Copilot's coding agent plus a control plane for running third-party agents on your repos. - [Goose](https://toolweight.com/tools/goose), Block's Apache-2.0 agent, MCP-native from the start, as both a CLI and a desktop app. - [Grok Build](https://toolweight.com/tools/grok-build), xAI's coding harness, launched July 2026, on this page but not yet verified. - [Kilo Code](https://toolweight.com/tools/kilo-code), Open-source VS Code agent in the Cline/Roo lineage, with an orchestrator mode and a CLI. ## Licence and attribution Data from Toolweight (https://toolweight.com), licensed CC-BY-4.0. - Licence: [CC-BY-4.0](https://creativecommons.org/licenses/by/4.0/) - Canonical HTML: https://toolweight.com/tools/diy-harness - Machine-readable: https://toolweight.com/tools/diy-harness.md · https://toolweight.com/api/v1 · https://toolweight.com/mcp - Toolweight takes no affiliate revenue and sells no placements.