Struct SetupReport 

pub struct SetupReport {
    pub login: StepResult<LoginInfo>,
    pub claim: StepResult<ClaimInfo>,
    pub claim_attempts: Vec<TierAttempt>,
    pub models: StepResult<ModelsInfo>,
    pub status: StepResult<StatusResponse>,
    pub auth_expired: bool,
}

Display-friendly summary of each step’s outcome. Returned by run so the caller can render however it wants (plain stdout, TUI body scrollback, future JSON output for scripting).

Fields

login: StepResult<LoginInfo>

claim: StepResult<ClaimInfo>

claim_attempts: Vec<TierAttempt>

Per-tier cascade history. Populated by step_claim with one entry per tier actually attempted (in cascade order Max → Pro → Lite). Empty when the cascade never ran (e.g. login failed upstream — claim is Skipped(CASCADE_FROM_UPSTREAM_FAIL)) or when a legacy test fixture wants the old single-row claim summary. render walks this to emit one row per tier so refused / errored intermediate tiers are visible, not hidden behind a single “claim failed” summary.

models: StepResult<ModelsInfo>

status: StepResult<StatusResponse>

auth_expired: bool

True when any API call rejected the stored bearer token (401/403). is_logged_in() only checks “does auth.toml exist” and get_valid_token only refreshes when the recorded expires_in says so — neither catches a server-side revocation or a refresh-token that the broker no longer accepts. Shells (TUI /codingplan, CLI atomcode codingplan) read this flag to drive an inline re-OAuth + retry instead of leaving the user staring at a “claim failed — run atomcode login again” line when /login would have fixed it in one step.

Implementations

impl SetupReport

pub fn render(&self) -> String

Render as a multi-line plain-text block for stdout / TUI body. Shared by the CLI subcommand and the /codingplan slash command so the visual contract stays consistent.

pub fn should_persist_config(&self) -> bool

True iff every persist-relevant step (login + claim + models) either succeeded outright or was skipped non-fatally (server reported duplicate=true, model list already current, etc.). Callers use this to decide whether to persist config changes to disk.

claim MUST be in the predicate: when claim returns Err (e.g. backend 500 like the AtomGit claim-v2 transaction- rollback bug), run() short-circuits and parks models as Skipped(CASCADE_FROM_UPSTREAM_FAIL) so the report stays focused on the actual failure. Without the claim check the gate flipped to true on every claim-failure path — triggering save_and_reload to rewrite config.toml unconditionally. That clobbered any manual edits the user made between TUI startup and /codingplan, and read as “claim failed but it still wrote models to my config”.

Trait Implementations

impl Clone for SetupReport

fn clone(&self) -> SetupReport

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for SetupReport

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.