Who Weftra is for

Weftra is not a copilot that sits beside you in your editor. It is a different way to work with coding agents: you hand off an outcome, an agent builds it inside Weftra's governed pipeline, and you get back a complete change — tested, reviewed, and on the record — to approve. You ask, it works, you review.

That shift, from typing beside an agent to delegating to one and reviewing the result, is the whole point. This page is about who that shift is for, and who it isn't for yet. For what Weftra does and does not do today, see What Weftra is (and isn't).

Weftra is for you if…

• You lead engineers who already delegate and review. You assign work and review pull requests today. Weftra adds a contributor you can hand a well-formed task to — one that always shows its work, runs the project's tests itself, and can never approve or merge its own change.
• You have to prove what AI changed. Security, compliance, or a regulator is asking who authorized this change, and can you show it? Weftra records the request, the spec that governed it, the agent and model that ran, the gates that passed, and who approved it — in a tamper-evident ledger on infrastructure you control.
• You look after many repositories. One spec can fan out across many projects, through one queue, with per-project policy and one view of status and cost — instead of a separate agent session per repository.
• You think in outcomes, not lines of code. Founders and product-minded builders can describe what they need in plain language. Weftra drafts the spec with you, asks about what's ambiguous, and brings back a reviewable change.
• You don't want to bet on one model vendor. Weftra governs the output, not which model wrote it. It is Claude-first today, with other engines reached through declared engine profiles.

For all of these, "ask, then review" isn't friction — it's the feature. You get leverage, because work keeps moving while you do something else, without giving up control, because nothing ships until a person you trust approves a change they can actually check.

Weftra is probably not for you (yet) if…

• You want faster typing in your editor. Inline autocomplete and chat beside your cursor are a different kind of tool, and a good one. Weftra complements it; it earns its keep when you hand off a whole unit of work, not when you want to speed up keystrokes.
• You want to skip review entirely. Weftra is built so a human stays accountable for what ships. If you want unattended merges with no oversight, that isn't what this is.
• You want a hosted service today. Weftra is self-hosted and single-tenant right now: you run it on infrastructure you control. A hosted option and packaged bring-your-own-cloud onboarding are on the roadmap, not shipped.

Why the wait is worth it

When you delegate a task you aren't waiting on a line of code. You are waiting on a finished, checked unit of work: a change with its tests run by Weftra rather than taken on the agent's word, a review by a reviewer that can block the change but never approve its own work, and a record of how it got there.

• Leverage: work is queued across tasks and projects and keeps moving while you're elsewhere.
• Trust: the change arrives with its evidence, so reviewing it is quick and confident rather than re-reading everything the agent did.
• Control where it matters: you approve the change, and the governance around it is what makes that approval mean something rather than a rubber stamp.

The one screen you live in: review

Because you're not beside the code, the place you are is the review. Weftra is built to make that moment good — a clear change, the evidence behind it, and governance you can rely on — so approving feels like getting work done for you, not checking a robot's homework.

Weftra is in active development. It says plainly what it does today and what is still maturing — see What Weftra is (and isn't), the capabilities overview, and the trust page.