Complete AI Training
Sign inGet my AI kit

Your job's AI kit

Get your AI kit

Tell us who you are and what you do. We show you your kit right away and email you the link: skills, prompts, AI agents, MCP servers and courses for your job.

500+ jobs ready, and we make a kit for any other job. No payment needed to look.

Share

AI agent for engineering managers

Technical Debt Prioritization Agent

Produce a ranked debt plan that fits capacity and supports the roadmap

Technical Debt Prioritization Agent: what goes in, what the agent does and what you get

What it does

The debt list has 150 items and the team works on whatever annoys them most. This agent reads the list, the incident history and the change frequency of the code involved. It scores each item by risk, such as incidents caused or likely, and by cost: how much work it takes and how much it slows other work. It builds a ranked plan that fits the team's capacity, usually 10 to 20% of each sprint. It checks the plan against the roadmap, spotting debt in areas that upcoming features will touch, and adjusts the order. It explains each ranking with the evidence. The manager approves the plan. Edge case: a small item blocks a big feature next quarter, so the agent moves it up.

How it works

Follow the arrows from top to bottom. The orange dashed arrow is the loop: when a check fails, the agent goes back and tries again.

Start and resultWhat it doesA check on its own workWaits for your OKGoes back and retries
Yes, continueApprovedNo 1 STARTS WHEN Quarterly planning 2 USES A TOOL Read the debt list and the incident history 3 USES A TOOL Measure change frequency of the affected code 4 DOES Score each item by risk and by cost 5 DOES Rank items by value per week of work 6 DOES Fit the ranked list to the team's capacity 7 USES A TOOL Compare the plan with the roadmap 8 CHECKS THE RESULT Do any roadmap features depend on debt that is notin the plan? If not: move that debt up and drop items at the capacityline. Back to step 5. 9 DOES Write the reasons for each item 10 YOU APPROVE Manager approves the plan 11 RESULT Ranked debt plan with evidence
Read the steps as a list
  1. Quarterly planning
  2. Read the debt list and the incident history
  3. Measure change frequency of the affected code
  4. Score each item by risk and by cost
  5. Rank items by value per week of work
  6. Fit the ranked list to the team's capacity
  7. Compare the plan with the roadmap
  8. Do any roadmap features depend on debt that is not in the plan?If not: move that debt up and drop items at the capacity line. Back to step 5.
  9. Write the reasons for each item
  10. Manager approves the planThe agent waits here for your OK.
  11. Ranked debt plan with evidence

How it decides

It ranks by risk reduction and speed-up per week of work, then moves items up when the roadmap depends on them and cuts the plan at capacity.

  • Score items higher when they caused incidents in the last 6 months
  • Score items higher in code changed more than 20 times a quarter
  • Reserve 15% of sprint capacity for debt by default
  • Move up any item that blocks a committed feature

Make it yours

Every agent is a starting point. You choose these settings for your own situation.

  • Capacity share (default 15%)
  • Scoring weights
  • Time window for incidents (default 6 months)
  • Roadmap source
  • Review schedule

What keeps you in control

It always asks you first

  • Manager approves the plan
  • Product owner agrees the capacity share

Hard limits

  • Never reorder committed roadmap items
  • Never close or delete debt items

It stops when

  • Done: the plan is approved and loaded into the tracker
  • Stop: the debt list has no evidence to score and needs triage first

Set it up

We guide you through the set-up, step by step

Members get the full set-up guide for this agent. No technical skills needed: you copy, paste and upload.

10 minto set it up in your AI
5 AIsChatGPT, Claude, Copilot, Gemini, Grok
  • One set of instructions to paste into your AI, with the clicks for ChatGPT, Claude, Microsoft 365 Copilot, Gemini and Grok
  • The agent then walks you through connecting your own data, one source at a time
  • A downloadable copy with the flow chart, the rules and the full guide
Get access to this agent

An example run

What happensThe list had 142 items. The agent scored all, and the top 12 fit a 15% capacity of 36 person-days. Item 31, a flaky queue library, ranked 9th. The roadmap check failed because the payments feature planned for next quarter depended on an old auth module not in the plan. The agent moved it to 4th and dropped item 12. The manager approved with a note on the change.

More agents for engineering managers