456
Prompt

Coding Agent Task Spec Prompt

Writes the task brief that keeps a coding agent inside scope.

1 min readupdated 2026-08-01

/ quick answer

Use to convert a loose request into a spec an agent can execute and you can verify. Writes the task brief that keeps a coding agent inside scope.

Writes the task brief that keeps a coding agent inside scope. Use to convert a loose request into a spec an agent can execute and you can verify. Copy the prompt below, swap the bracketed variables for your own context, and run it in any capable model. This prompt node is part of the Onexial knowledge graph and links to related concepts, workflows and tools below.
Context
Use to convert a loose request into a spec an agent can execute and you can verify.
Prompt
Write a task specification for a coding agent working in an existing repository. Include:

1. GOAL — one sentence, outcome not implementation.
2. ACCEPTANCE CRITERIA — checkable statements, including the command that must pass.
3. SCOPE — files and modules it may change; explicitly list what it must NOT touch.
4. CONTEXT — existing patterns, types and conventions it must follow.
5. NON-GOALS — refactors and improvements it should deliberately skip.
6. VERIFICATION — the exact command to run and what output proves success.
7. RISKS — what could break elsewhere and how to check.

Keep the diff under 400 lines. If the task is larger, split it into ordered sub-tasks.

Request: {{REQUEST}}
Stack: {{STACK}}
Repo conventions: {{CONVENTIONS}}
Example Output
1. GOAL — Users can filter the orders table by status without a page reload. 2. ACCEPTANCE — `bun test` passes...
Related Workflow
Related Tool Stacks
/ frequently asked

What does the Coding Agent Task Spec Prompt prompt do?

Use to convert a loose request into a spec an agent can execute and you can verify.

Which AI models work with this prompt?

It is model-agnostic: it works with any capable general model. Replace the bracketed variables with your own context before running it.

What output should I expect?

1. GOAL — Users can filter the orders table by status without a page reload. 2. ACCEPTANCE — `bun test` passes...

/ topics#prompt#ai-development#coding
↳ connected nodes
Workflow↳ linked
Run a Coding Agent on a Real Codebase Safely
Give an agent write access without letting it wreck main.
Tool Stack↳ linked
Coding Agent Stack
Run coding agents with executable feedback and reviewable diffs.
Dictionary↳ linked
Coding Agent
A coding agent reads a repository, plans a change, edits files, runs tests and iterates until the task passes — instead of just suggesting snippets.
Dictionary↳ linked
AI Software Engineering
AI software engineering is the practice of building software where agents write most of the code and humans own architecture, review and verification.
Dictionary↳ linked
Context Engineering
Context engineering is the discipline of deciding exactly what information enters a model's context window, in what order and at what cost.
Comparison↳ linked
Coding Agent vs Autocomplete Copilot
Copilots accelerate typing; coding agents complete tasks. Different unit of work, different review burden.