A ready-to-use system prompt for an issue triage agent that first tries to answer from your docs, and only when a real bug remains, drafts a clean, reproducible GitHub issue — with human approval before it's opened. Copy it, fill the brackets, and paste it into your agent's instructions.
What this agent does
Most "issues" are really questions your docs already answer. This prompt makes the agent answer those first from your documentation and repo, deflecting noise from your tracker. When something is a genuine bug, it gathers the details that make an issue actionable — version, steps, expected versus actual — and drafts the issue. Through MCP it can open the issue, but only behind human write-confirm, so nothing lands in your tracker without a maintainer's approval.
The prompt
You are the triage assistant for [Project]. First, try to answer the
user's problem from the provided context drawn from [Project]'s docs and
repo, and cite the source. Only if it's a genuine bug, collect version,
environment, reproduction steps, and expected vs. actual behavior, then
draft a concise GitHub issue. Do not open the issue without explicit
confirmation. Never invent reproduction details or fabricate fixes.
How to use it
Replace [Project] with your project name, index your docs and repo, connect your issue tracker over MCP, and paste the prompt into your agent's instructions. Writes pause for approval automatically. See the GitHub MCP integration and the open source use case, then start free.
Why grounding beats a bare prompt
A prompt alone doesn't make a triage agent reliable — grounding does. Pasted into a generic chatbot, these instructions still let the model answer from its general knowledge and invent details. In AIML.chat the same prompt sits on a retrieval layer that feeds the agent only the content you indexed, so "answer only from the provided context" is enforced by the system rather than merely requested. Every claim links back to a page you control, and the analytics record what people ask so you can refine both the prompt and the content behind it. That pairing — your instructions plus your own grounded content — is what turns this template into a dependable agent.
Will it open issues on its own?+
No — it drafts the issue, but opening it requires human approval through AIML.chat's write-confirm flow.
Does it deflect non-bugs?+
Yes — it answers questions from your docs first and only triages when a genuine bug remains.
Is there a free plan?+
Yes — 100 messages a month and one website on the free tier. See pricing for higher limits.