What counts as evidence?
An achievement starts as a claim. It can be manually sourced or connected to an exact, uniquely matched quotation in a user-provided source document. A quotation shows where a statement came from; it does not independently prove that the underlying statement is true.
You decide whether to confirm, reject or leave each claim unverified. Only participant-confirmed achievements can support manual job-requirement links or enter optional recovery AI-review payloads.
What happens when it changes?
Correcting an achievement records its history, resets its confirmation and revokes dependent requirement links. This keeps an old decision from silently supporting a new statement.
Missing evidence stays visible. The app does not fill a gap by inventing a career achievement or treating a keyword match as a verified qualification.
Three routes. Three hypotheses.
Similar work to before
Test current role requirements against your confirmed achievements. Hiring availability still needs to be checked.
A new kind of work
Identify transferable evidence and a specialty-specific gap. Prior experience does not automatically establish readiness.
A stepping-stone role
Define a narrow deliverable and test interest. Demand, timing and actual income remain unknown.
Where AI fits.
Core recovery planning is deterministic and works without an AI provider. Optional Anthropic review requires local configuration, an exact JSON payload preview and explicit consent for each request. Suggestions do not automatically change your session. Workspace scoring and AI tools are separate workflows; their scores are not recovery-route predictions or hiring probabilities.
Read the recovery AI guide ↗Your next move starts locally.
Open source. Optional AI. You set the direction.