Audit Api Bundle Laravel Package

Getting Started

• Install the package with Composer, publish its config and migrations, then run php artisan migrate.
• Look first at the package config (auditable models, storage connection, retention settings) and at php artisan audit:logs --help to see available filters.
• Start with a single model: add the package’s auditing trait/interface, perform a create/update/delete, then run php artisan audit:logs and confirm the Changes column shows a readable diff.
• In v0.9.12 the Changes column is pretty-printed JSON, so the command is now a practical first-stop debugging tool for “what changed and when” without opening the database or dumping raw JSON.

Implementation Patterns

• Use audit:logs as the default inspection workflow: run it after reproducing a bug or deploying, filter by model/date/user where supported, and read the Changes column directly.
• For deep investigation, redirect output to a file and pipe it through jq or your log tooling; the pretty-printed diffs are easier to scan but still benefit from structured post-processing.
• Keep auditing enabled on models that carry business-critical state, and use the command as a lightweight regression check in staging after seeders or migrations.
• When reviewing a change set, compare the Changes column against the event type (created/updated/deleted) and the audit timestamp; the diff is display-only in v0.9.12, so the underlying stored payload remains the source of truth.
• If the package supports custom audit resolvers or mutators, keep diff formatting concerns out of them; let the command handle presentation so exports and API consumers still receive raw JSON.

Gotchas and Tips

• Pretty-printed JSON can wrap or truncate in narrow terminals; use a wider terminal, redirect to a file, or disable ANSI output when copying diffs into tickets.
• Large updates produce large diffs; filter the command output by model, record ID, or date range before inspecting, otherwise the Changes column becomes noisy.
• The v0.9.12 change affects display only. Existing audit rows will render pretty-printed automatically, but stored JSON is unchanged—do not expect new columns or schema changes.
• If Changes is empty, verify the model actually fires audit events, the trait/interface is applied, and the audit driver is configured; the command can only show what was recorded.
• Check JSON casting on the audit model if diffs appear as escaped strings rather than structured JSON; the pretty-printer can only format valid decoded JSON.
• Use the command’s exit behavior carefully in scripts: it is an inspection tool, not a validation command, so assert on database state or exported JSON rather than parsing human-readable output.