CLI error message when AppHost exits before backchannel connection is vague and missing exit code

[ellahathaway]

Is there an existing issue for this?

• I have searched the existing issues

Describe the bug

When the AppHost process exits before the CLI can establish a backchannel connection, the error message is:

An error occurred while connecting to the AppHost. The AppHost possibly crashed before it was available: AppHost process has exited unexpectedly..

Problems:

1. No exit code — the user has no idea why the process failed
2. Redundant phrasing — possibly crashed + exited unexpectedly say the same thing
3. Double period at the end — the inner message ends with . and the format string also appends .

Expected Behavior

The error should include the exit code and clearly describe what happened:

An error occurred while connecting to the AppHost: The AppHost process exited unexpectedly with exit code 42.

Steps To Reproduce

1. Have an AppHost project that crashes on startup (e.g., build error, missing dependency)
2. Run aspire run
3. Observe the vague error message

Anything else?

This affects both DotNetCliRunner (local) and GuestAppHostProject (remote/server) paths.

[joperezr]

🤖 Aspire triage pass — 2026-05-19 22:33 UTC

Area: area-cli
Kind: bug
Repro: yes-minimal — see "Steps To Reproduce" section of #17277
Affected version(s): not specified (latest: 13.3.x train)
Severity: S4 — error message quality / polish; user can still inspect AppHost logs to determine cause

Urgency: ⚪ backlog — no milestone
Reason: S4 polish issue (error wording + missing exit code in CLI message); no broken core scenario, workaround exists (read AppHost logs directly)

Suggested owner(s): @JamesNK (dominant area-cli author last 90d), backup @mitchdenny
Evidence:

• @JamesNK merged 70+ PRs in area-cli in the last 90 days (e.g. Move InitCommand hardcoded strings to resource file for localization #17260, Improve ConsoleCancellationManager with process termination timeout #17251, Route error-context DisplayMessage calls to stderr in BaseCommand #17231, Use grid layout in DisplayMessage/DisplayError for better text wrapping #17198, Fix flaky NewCommandChannelResolutionTests by isolating home directory #17180 within the last 3 days)
• @mitchdenny merged Fix config discovery to search from apphost directory and add aspire.config.json to .NET templates #15423, Remove mature feature flags that no longer need toggles #15075, Fix CopilotCliRunner version parsing for 'GitHub Copilot CLI X.X.X' format #14568, Fix TryGetResourceToolMap cache miss causing perpetual tools/list refresh loop #14539 in area-cli
• No specific file paths cited in the issue body to score CODEOWNERS against; DotNetCliRunner and GuestAppHostProject referenced are in the CLI source tree

Confidence: medium
Needs human?: no — straightforward CLI polish; owner can decide whether to slot into 13.4 or leave in backlog

Dry-run mode — no labels or milestone applied. Reply /triage apply to promote, or edit and reapply.

[ellahathaway]

Semi-related to #17244

3. Localization: "Waiting for dependencies: {0}." should not end with a period
File: src/Aspire.Dashboard/Resources/Columns.resx (and all xlf files)

The string "Waiting for dependencies: {0}." places a period after the interpolated dependency list, which reads awkwardly (e.g., "Waiting for dependencies: nginx, redis."). Messages ending with : followed by content should not include a trailing period.