Part
Coding agent
You might say: agent, AI coding agent, model plus harness
A model harnessed with tools, a system prompt, and a context window, configured to carry one coding task through a run.
- What it does
- Turns a task into model decisions, tool calls, and a result by letting the harness repeat the agent loop.
- Ask
- What model and harness should drive this task, and what must the run leave behind?
What to look for
- A run page names the model and harness
- A tool trace shows model turns and tool results alternating
- A final result links to changed files or verification evidence
Parts
- Agent harness: The program around a model that drives an agent loop in a run.
- Model choice: The policy that routes tasks to different model categories or tiers by difficulty and cost.
- Effort: A dial for how much reasoning the model does before it answers.
- Context window: Everything the model sees on each model provider request.
- Permission mode: The setting that decides which tool calls run automatically and which stop for a permission request.
Connections
- See also: Instructions
- See also: Tool access
- See also: Security boundary
Sources
- Matt Pocock, AI Coding Dictionary, Agent
A model harnessed with tools, a system prompt, and a context window, that takes turns with a user. The model in motion.
- Matt Pocock, AI Coding Dictionary, Model
The parameters. Stateless — does next-token prediction and nothing else. Cannot do anything agentic on its own.
- Matt Pocock, AI Coding Dictionary, Model
Everything that feels like an agent working — choosing tools, reading results, looping until the task is done — is the harness orchestrating many of those predictions in a row.
- Anthropic, Building effective AI agents, Agents
Once the task is clear, agents plan and operate independently, potentially returning to the human for further information or judgement.
- Anthropic, Building effective AI agents, Agents
During execution, it's crucial for the agents to gain “ground truth” from the environment at each step (such as tool call results or code execution) to assess its progress.
Use it on your project
teardown-apply: hold "Coding agent" (coding-agent) from the teardown at https://teardownatlas.com/software-factory/ against my project
Paste it into a coding agent that has the teardown skills.
An AI agent wrote this map with the teardown skill. It can contain errors: check the sources before you rely on a part. About and legal