[Hackathon] Add "test open <id>" / "--web" to jump from the terminal to the dashboard

[Andy00L]

Discord Username / User ID

interferon0

What does this improvement do?

The CLI already computes a dashboard deep-link (resolvePortalUrl in src/lib/facade.ts, attached as dashboardUrl to run output and meta.portalUrl on history) but only ever prints it as text — there is zero browser-opening code in src/. Add test open plus a --web flag on test result, test run --wait, test rerun --wait, and test failure summary that opens the OS browser at that URL, with a --no-browser companion that only prints the URL (SSH/headless/CI/agents). This is the gh browse / cypress open pattern: a failing run's video/DOM/timeline lives in the dashboard, and today reaching it means copy-pasting a URL the CLI already built.

Details / implementation notes

Spawn open / xdg-open / start via node:child_process with an args array (never a shell string). When resolvePortalUrl returns undefined (unknown host), print a clear "no dashboard for this endpoint" line and exit non-zero. The URL builder already refuses userinfo and non-https production hosts, so the safety half is done. No new dependency. Tests: injected spawner per platform, --no-browser prints only, unknown-host error path. PR to follow once assigned.

Confirmations

• I have searched existing issues and this is not a duplicate.
• I have provided my Discord identity above for reward coordination.

[zeshi-du]

Triage: accepted — the dashboard URL is already computed; opening it is the missing last step. args-array spawn (never a shell string) and --no-browser as proposed. Comment /assign to claim.

[Andy00L]

PR ready here: #182

[Andy00L]

/assign

[testsprite-hob]

Assigned to @Andy00L. Thanks for taking this on. ⚠️ You're now over the soft 3-issue limit (9/3). Consider finishing one before taking more.

Your open assigned issues (9/3):

• [Hackathon] "test steps --run-id" silently discards the per-step error text the API already returned #126 — [Hackathon] "test steps --run-id" silently discards the per-step error text the API already returned
• [Hackathon] Add "test diff <runA> <runB>" to isolate what regressed between two runs #124 — [Hackathon] Add "test diff " to isolate what regressed between two runs
• [Hackathon] Make "test wait" variadic: wait on several run-ids in one invocation #120 — [Hackathon] Make "test wait" variadic: wait on several run-ids in one invocation
• [Hackathon] Honor HTTPS_PROXY / HTTP_PROXY / NO_PROXY for corporate and CI networks #119 — [Hackathon] Honor HTTPS_PROXY / HTTP_PROXY / NO_PROXY for corporate and CI networks
• [Hackathon] Add "testsprite test scaffold" to emit a schema-correct starter plan #101 — [Hackathon] Add "testsprite test scaffold" to emit a schema-correct starter plan
• [Hackathon] Read ~/.testsprite/config for persistent defaults (output, endpoint, project) #100 — [Hackathon] Read ~/.testsprite/config for persistent defaults (output, endpoint, project)
• [Hackathon] Add "testsprite test lint" to validate plans/steps offline before creating #98 — [Hackathon] Add "testsprite test lint" to validate plans/steps offline before creating
• [Hackathon] TESTSPRITE_PROJECT_ID env default for --project #76 — [Hackathon] TESTSPRITE_PROJECT_ID env default for --project