litlaunch inspect checks local runtime readiness without launching Streamlit,
opening browsers, starting a diagnostics server, or dumping raw environment
variables.
Human-Readable HTML Report
litlaunch report app.py
litlaunch report --profile my-webapp
litlaunch report --profile my-webapp --output my-report.html --force
litlaunch report --profile my-webapp --open
litlaunch report is the recommended human-facing diagnostics workflow. It
writes .litlaunch/reports/litlaunch-report.html unless --output is
provided. Existing report files are not overwritten unless --force is
supplied. --open opens the generated standalone HTML report in the default
browser after writing it; if opening fails, the report still generates and
LitLaunch emits a warning.
The explicit inspect form remains available:
litlaunch inspect app.py --html
litlaunch inspect app.py --html --output litlaunch-report.html
litlaunch inspect --profile my-webapp --html --output litlaunch-report.html
HTML output is the recommended human-readable diagnostics artifact. It is a standalone, dependency-free report meant to be readable at a glance, printable, and shareable after review. It includes LitLaunch version, platform capability, Streamlit availability, browser capability/resolution, target launch-plan details, and profile runtime settings when a profile is loaded. It does not start a local server, run JavaScript, load external CSS, inspect browser URLs, launch Streamlit, or open a browser.
With a target app, the report adds:
- app path existence
- runtime governance summary with allowed/blocked posture, highest severity, primary finding, and top recommendation
- host binding and network exposure posture
- runtime exposure scope, active trust mode, acknowledgement state, and whether the current trust mode allows the configured binding
- transport security posture, including Streamlit-native TLS detection, incomplete TLS configuration, and network-visible plaintext HTTP risk
- operational posture reminders for loopback shutdown hooks, browser ownership, diagnostics privacy, and plaintext profile environment values when present
- command display
- app URL display
- health URL display
- browser resolution summary
For wildcard bind hosts such as 0.0.0.0 and ::, URL displays use the local
client URL LitLaunch can connect to for browser launch and health checks, while
the host binding/posture items still show the configured network-visible bind
address.
The posture sections are operational, not a compliance rating or policy engine.
They help identify local-only versus network-visible runtime configuration, but
LitLaunch does not authenticate users, terminate TLS, or secure the Streamlit
app itself.
Streamlit-native TLS settings can be detected when supplied through
streamlit_flags, but certificate paths are summarized rather than included in
diagnostics.
Plain litlaunch inspect prints concise guidance for choosing HTML, JSON, or
support bundle output. It no longer renders a full legacy text report.
JSON
litlaunch inspect app.py --json
litlaunch inspect app.py --json --output litlaunch-report.json
JSON is for tools and automation. It uses the same structured report model as HTML and bundle output.
Sanitized Bundle
litlaunch inspect app.py --bundle
litlaunch inspect app.py --bundle --output litlaunch-report.txt
The support bundle is concise and copyable for issues or support requests. It carries the same privacy warning as the HTML report: sanitization is pattern-based, avoids raw environment dumps, and still requires manual review before sharing.
Sanitization
Inspect output avoids:
- shutdown tokens
- raw environment variable dumps
- full PATH dumps
- sensitive-looking values such as token, secret, password, and key values
- common local home/user path prefixes where practical
Existing output files are not overwritten unless --force is supplied.
Sanitization is pattern-based and intentionally lightweight. It is appropriate for LitLaunch's local diagnostics workflow, but it is not a cryptographic scrubber. Encoded, base64, URL-wrapped, or heavily reformatted secrets may not always be detected. Review support bundles before sharing them publicly.
Dashboard Scope
There is no local diagnostics dashboard or diagnostics web server today.
The HTML report is a standalone file generated by litlaunch report. It
summarizes platform, Streamlit, browser, target/profile, launch-plan, and
sanitization details in one shareable artifact without running a local web
server or loading external resources.