Technical-debt remediation plan
Turn a vague cleanup request into evidence-ranked debt with bounded, verifiable work packages.
Read this prompt as Markdown or plain text
When to use it
The evidence and cost-of-delay requirements keep the plan tied to product risk rather than aesthetic preference, while work packages make the result executable.
Prompt
Assess technical debt in [SCOPE] and produce a remediation plan. Do not edit code. Build an evidence inventory from code, tests, build output, dependency reports, incidents, performance data, and existing TODOs. Do not label unfamiliar code as debt merely because you prefer another style. For each candidate debt item record: - exact location and current behavior - evidence that it creates defects, slows delivery, increases security/operational risk, or blocks a stated goal - impact and likelihood - dependencies and affected owners/surfaces - estimated remediation size: S / M / L / XL - cost of delay and what happens if it remains unchanged Rank items with a stated method using impact, urgency, confidence, and effort. Group the selected items into independently shippable work packages. Each package must include scope, non-goals, prerequisites, implementation outline, migration/rollback concerns, acceptance criteria, verification commands, and observability needed after release. Separate: 1. Fix now 2. Schedule with a trigger or deadline 3. Monitor with a named signal 4. Accept intentionally, with rationale End with the top three recommended actions, the evidence behind their order, quick wins, high-risk unknowns, and the next review date.
Adapt it to your task
Condensed and made tool-neutral from GitHub Awesome Copilot's Tech Debt Remediation Plan agent.
Third-party prompt. Source review does not establish benchmark performance or guarantee an outcome.
Source and license
Contributor: @aaronpowell. Repository: github/awesome-copilot.
License: MIT. Pinned commit: b95e24caae4f25c5985f8527fd092ccfed039c36.
Review: source-reviewed, . Compared with the commit-pinned source and adapted into a tool-neutral, copy/paste request. This is a non-visual task, so no output artifact is required.