[thebrandonlucas]

Stack position: 7 of 13 in the look-and-feel stack. Base: feat/cli-look-06-brand (#105), whose brand color this uses.

Problem

Signing in is the first moment a person spends with the CLI, and it ended on a plain line. Nothing on the terminal said "Voltage".

Behavior

After a successful voltage login on an interactive stderr:

    __              ____
   / /   _  _____  / / /____ ____ ____
  / /_  | |/ / _ \/ / __/ _ `/ _ `/ -_)
 /_  /  |___/\___/_/\__/\_,_/\_, /\__/
   //                       /___/

✓ Logged in as you@example.com (account: me)

• A lightning bolt in bold yellow beside the name in figlet's smslant font. ASCII only, so every font draws it; five lines, 39 columns wide.
• The name is in Voltage orange at the terminal's depth (24-bit, 256-color, or bold), from feat(cli): color help headers Voltage orange and honor FORCE_COLOR #105. The bolt uses a Bolt role that is decoration only, never a status.
• Left out on a terminal narrower than 40 columns (it would wrap), with --quiet, and when stderr is redirected. Never on stdout.
• PR 13 reuses it for the greeting on bare voltage.

Try it on staging

Setup: Try a pull request on staging (added by #97).

voltage login --account me --credential-store file

Expect: after you approve in the browser, the yellow bolt and orange wordmark, then ✓ Logged in as ….

COLUMNS=39 voltage login --account me --credential-store file
voltage login --account me --credential-store file -q
voltage login --account me --credential-store file 2>&1 | cat

Expect: no wordmark in any of them. The first still shows ✓ Logged in as. The second hides that line too. The third prints the old plain lines.

Validation

• cargo fmt --check, cargo clippy --locked --all-targets -- -D warnings, cargo machete --with-metadata, cargo test --locked (158 unit and 126 integration tests), and python3 scripts/check-coverage.py all pass.
• Unit tests check that the bolt and name are ASCII, that the bolt lines share one width, that the name has no trailing spaces, and that both fit the width. They also check the narrow-terminal rule (unknown width shows it, 40 shows it, 39 leaves it out), and that each line is a yellow bolt then the name in the 256-color brand color, followed by a blank line.
• The device-login PTY test now checks that the bolt and wordmark sit directly above ✓ Logged in as, and that the first line has the exact bolt and brand escape sequences.
• CRAP: 0 of 717 measured functions at or above 20. The highest is 18.00 (src/cli.rs endpoint, untouched).

Docs

docs/guides/login-and-profiles.md and CHANGELOG.md.

[tee8z]

Approved 01ff3ef against feat/cli-look-06-brand (#105 80f0dec). The wordmark goes only to stderr, and only when stderr is a terminal. It is hidden with --quiet through notice_line. It is 39 ASCII columns, and the 40-column fits gate handles COLUMNS. Only static text is drawn, and anstream strips the color with NO_COLOR or TERM=dumb.

Nonblocking: the CHANGELOG line, "a lightning bolt beside the name in Voltage orange", reads as if the bolt is orange, but the code paints it yellow (Role::Bolt), as the guide correctly says. No terminal test covers --quiet hiding the wordmark.

Validation on this head, in nix develop: cargo fmt --check, cargo clippy --locked --all-targets -- -D warnings, cargo machete --with-metadata, cargo test --locked (158 unit, 126 integration) and python3 scripts/check-coverage.py passed. The CRAP report did not finish: its instrumented run hit the known flaky #75 terminal test (a_profile_saved_after_login_keeps_changes_made_while_its_prompt_was_open), which #104 addresses. CI has not run on this head. The stack also needs a rebase onto master.

[tee8z]

Re-approved 4e764c3, a rebase onto #105 73ba6b3. The code matches 01ff3ef, which I approved; only the login guide's sentence changed, picking up #96's countdown wording. Validation on this head, in nix develop: cargo fmt --check, cargo clippy --locked --all-targets -- -D warnings, cargo machete --with-metadata, cargo test --locked (172 unit, 133 integration) and python3 scripts/check-coverage.py passed. CRAP: 750 measured functions, none at or above 20, highest wait_for_payment at 17.05. CRAP passed on its second run, after a pseudo-terminal timeout on the first (#104). CI is green. This also closes the gap from my earlier approval, which had no finished CRAP report.