From rolepod-wplab
Multi-probe diagnostic + security audit of a connected target — plugin conflicts, slow queries, large autoload options, broken images, PHP errors, outdated core/plugins/themes, weak admins. Phase = Debug.
How this skill is triggered — by the user, by Claude, or both
Slash command
/rolepod-wplab:wp-diagnoseWhen to use
user asks "what is wrong with this site / audit security / check for problems / why is the site slow", OR after a regression spike, OR before a migration to a new environment
The summary Claude sees in its skill listing — used to decide when to auto-load this skill
If the marker file `$GIT_ROOT/.rolepod/parent-active` exists, follow the
If the marker file $GIT_ROOT/.rolepod/parent-active exists, follow the
with-rolepod mode declared in the frontmatter — produce reduced output
suited to the parent's debug-issue phase orchestrator. Skip the fix-flow
narrative; emit WP-runtime findings only (error log lines, hook trace, query
log) and let the parent decide remediation.
Otherwise, follow standalone mode — run the full diagnostic flow described below.
GIT_ROOT=$(git rev-parse --show-toplevel 2>/dev/null) || GIT_ROOT="$PWD"
if [ -f "$GIT_ROOT/.rolepod/parent-active" ]; then
MODE=with-rolepod
else
MODE=standalone
fi
Heavyweight probe + audit. Multi-second, ranks findings by severity, writes a markdown / JSON artifact. Subsumes audit_security + audit_many + cron/cache/mail/user_session inspection. The "what is wrong here" entry point — wp-health-check answers "is it up", this answers "is it healthy at depth".
audit_many orchestrator.Skip when:
wp-health-check.wp_options for plugin X?" → wp-introspect.Owns:
rolepod_wp_diagnose (5 scopes: plugin_conflict_probe / slow_queries / large_options / broken_images / php_errors).rolepod_wp_audit_security (core/plugin/theme outdated + weak admins + WP_DEBUG flag).rolepod_wp_audit_many (parallel audit across N target_ids).rolepod_wp_cron_tool op=list (inspect schedule).rolepod_wp_cache_tool op=inspect (object-cache type + transient counts).rolepod_wp_mail_test (SMTP / wp_mail probe).rolepod_wp_user_session_list (active sessions for security view).Does not own:
wp-content, wp-edit-plugin, wp-execute-php, wp-migrate).wp-introspect.wp-health-check.Return / hand off:
wp-content to update via REST.wp-edit-plugin or direct option_set.audit_many for the consolidated index.scopes array (default = all except broken_images, which is expensive).report_format (markdown / json, default markdown).| User intent | Scopes |
|---|---|
| "security audit" | audit_security + user_session_list |
| "site is slow" | diagnose (slow_queries + large_options) + cache_tool inspect |
| "plugin conflicts" | diagnose (plugin_conflict_probe + php_errors) |
| "before migration" | audit_security + diagnose (all scopes) + cron_tool list |
| "full sweep" | every scope above |
| "audit 20 sites" | audit_many |
Most run in parallel (audit_many always does). For single-target multi-scope: sequential to keep companion side cool.
Severities: CRITICAL > HIGH > MEDIUM > LOW > INFO. The per-probe handler returns this already; the skill aggregates + sorts.
./.rolepod-wplab/artifacts/<run_id>/diagnose-report.{md,json} — per templates/diagnose-report.md.
Print TOP 5 findings (sorted by severity) + path to artifact. Do not dump all findings inline — that defeats the artifact's purpose.
rolepod:security-engineer for the audit_security pass interpretation.rolepod:performance-engineer for slow_queries + large_options + cache findings.rolepod:qa-tester for the full sweep + ticket creation.Diagnose report — templates/diagnose-report.md. Canonical handoff format.
Read when picking scopes for an unfamiliar site type or when interpreting a non-trivial finding:
examples/diagnose-examples.md — good vs bad scope pick for "site is slow"; good vs bad reading of audit_security output.Inline only. Each probe's contract is documented in src/tools/composite/wp_diagnose.ts + wp_audit_security.ts. The artifact format is the template.
Full Rolepod adds historical trending: today's diagnose vs yesterday's, surfaces regressions explicitly. Standalone, the user compares manually via the artifact path.
wp-content for option fixes, wp-edit-plugin for plugin config, wp-execute-php for runtime fixes, wp-migrate for "let's move to a cleaner host".audit_many index.wp-health-check periodically going forward.Guides creation, editing, and verification of skills for AI coding agents using test-driven development with subagent scenarios. Use when authoring or debugging skills.
npx claudepluginhub nuttaruj/rolepod-wplab --plugin rolepod-wplab