Improve lookup performance, submission activity, and form building

[Shorton88]

ActionStack 0.5.4 improves lookup response time, submission visibility, and the form builder.

• Simple KV Store lookups with declared string columns query the collection directly using the requesting user's permissions. Transformations, filtered definitions, and multivalue results retain SPL execution. Lookup requests skip repeated setup migrations and occupied rate-limit slots.
• Multiple-value fields accept comma-, newline-, and semicolon-separated lists. Lookup verification ignores case, removes case-only duplicates, and preserves stored values. Suggestions stay inside the form and are cached per field for 30 seconds; submission always revalidates.
• Submissions show playbook/action totals and status counts on ten-row pages. Only the current page is polled, with three reads in flight. Receipts prefer custom action names and display reported block results, summaries, and collapsed result data.
• The builder adds conditional static text, grouped collapsible field controls, and Scan, Server, and User icons. Scan uses a targeting reticle with a heartbeat line. Automation starts enabled for new forms; existing and cloned forms keep their setting. Navigation remains visible during content scrolling, and avatars retain circular dimensions.
• Builder labels and inputs use 14px text, with 13px help text and wider properties panels. Submission table and receipt text are also larger. Lookup search textareas have room for four lines.
• Remove unsupported inputlookup local options and correct the guidance. Existing searches with local=true or local=false receive a correction message; remove the option and republish those forms.

Validation: 33 JavaScript and 177 Python tests pass; production build and the 0.5.4 Splunk package succeed. Browser checks cover mixed-case CSV normalization, preview validation, builder disclosures, pagination, custom names, status counts, expandable result data, both themes, sidebar scrolling, larger lookup text at desktop and narrow widths, and the Scan icon. Tests use simulated services; live Splunk lookup latency and SOAR behavior remain unverified.

Block coverage depends on SOAR's returned data. Format/filter/decision/code datapaths are read for the latest three playbook runs. Utility rows require explicit custom-function run headers in the playbook report; missing statuses remain Unknown. Run totals may cover more history than the bounded status-count window, which is marked in the UI.