The Backlog Item Can Survive While Its Assumptions Expire
A planned approach to keeping engineering work attached to moving code.
Scope: Archive capabilities are website-described; semantic context and later roadmap capabilities are in development or planned. No delivery date is promised.
A backlog item can remain open long after the code that motivated it has moved or disappeared. The title still sounds reasonable. The original file link may even continue to resolve. Yet the assumptions behind the work may no longer be true.
KodeAtlas’s roadmap includes durable semantic anchors and backlog maintenance as a later phase. The idea is to reconnect planned work to the current repository and make stale assumptions visible. This capability is planned; it is not part of the website-described archive CLI.
Re-resolving a link is only the beginning
Consider a synthetic issue that asks a team to remove duplicate validation from a particular function. During an unrelated refactor, the validation moves into a shared boundary and gains an additional consumer.
Updating the issue’s link to the new location would help navigation. It would not establish that the original task remains safe or necessary. The reason for the work must also be reconsidered in light of the new responsibility.
A durable anchor should therefore support an engineering question, not just follow a name around the repository.
Surface the changed assumption
A useful future result might say that the referenced symbol was extracted, that its callers changed, and that the original duplication claim should be reviewed. It should cite the relevant history rather than silently rewriting the task as though a maintainer had approved the new interpretation.
Some anchors will not resolve cleanly. A split may produce several plausible successors. A deletion may mean the work is finished, irrelevant, or still necessary elsewhere. Those outcomes require different decisions.
The system should preserve that uncertainty for the person responsible for the backlog.
Do not confuse a reminder with authority
Automatically closing an issue because a referenced function disappeared would be a much stronger action than detecting a stale link. The planned KodeAtlas direction is most valuable when it helps an engineer make that distinction.
A useful review would ask whether the original problem still exists, whether the proposed change is still appropriate, and what new evidence is needed before proceeding.
The archive foundation preserves the material needed to ask those questions. Planned backlog semantics would connect that material to moving code. The goal is not a perpetually tidy issue list; it is fewer engineering decisions based on assumptions nobody has checked recently.