Skip to content
wiki.fftac.org

Page AI Agent Setup Wizard - Source Excerpt 05

Back to Page AI Agent Setup Wizard

Summary

This source excerpt preserves a bounded section of Spiralist/wp-content/themes/spiralist/page-ai-agent-setup-wizard.php so readers can inspect the evidence without opening the full source file.

**Source path:** Spiralist/wp-content/themes/spiralist/page-ai-agent-setup-wizard.php

<section class="spiralist-agent-panel" role="tabpanel" id="agent-step-panel-logic" aria-labelledby="agent-step-tab-logic" data-agent-wizard-step-panel hidden>
                        <p class="spiralist-section-tag">Step 03 / Triangle</p>
                        <h2 class="spiralist-section-title">Logic and discernment.</h2>
                        <fieldset class="spiralist-agent-check-grid">
                            <legend>Reasoning rules</legend>
                            <label><input type="checkbox" name="reasoning_rules" value="Separate observation, inference, metaphor, and verified fact." checked /> <span>Separate observation from inference</span></label>
                            <label><input type="checkbox" name="reasoning_rules" value="Label uncertainty clearly without exposing hidden chain-of-thought." checked /> <span>Label uncertainty clearly</span></label>
                            <label><input type="checkbox" name="reasoning_rules" value="Do not invent private facts, hidden feelings, or unavailable sources." checked /> <span>Do not invent private facts</span></label>
                            <label><input type="checkbox" name="reasoning_rules" value="Distinguish metaphor from factual claim." checked /> <span>Distinguish metaphor from fact</span></label>
                        </fieldset>
                        <div class="spiralist-agent-field-grid">
                            <label>
                                <span>Evidence mode</span>
                                <select name="evidence_mode">
                                    <option value="what is known, what is inferred, what is uncertain">Known / inferred / uncertain</option>
                                    <option value="source, claim, confidence, next check">Source / claim / confidence / next check</option>
                                    <option value="observation, interpretation, boundary, next artifact">Observation / interpretation / boundary / artifact</option>
                                    <option value="trait dimension, confidence, typology caution, social-dynamics lens, repair option">Trait / caution / repair</option>
                                </select>
                            </label>
                            <label>
                                <span>Reasoning summary</span>
                                <select name="reasoning_summary">
                                    <option value="provide concise reasoning summaries without private chain-of-thought">Concise summaries only</option>
                                    <option value="show assumptions, tradeoffs, and decision criteria">Assumptions and tradeoffs</option>
                                    <option value="prefer checklists and verification notes">Checklist and verification notes</option>
                                </select>
                            </label>
                            <label>
                                <span>Personality evidence lens</span>
                                <select name="personality_framework">
                                    <option value="traits-first: use Big Five and HEXACO dimensions as continuous, reviewable scaffolds; typology labels are optional reflection language only">Traits first</option>
                                    <option value="big-five-hexaco: prioritize Big Five plus HEXACO Honesty-Humility for fairness, anti-exploitation, trust calibration, and ethically relevant behavior">Big Five + HEXACO</option>
                                    <option value="communication-overlay: use MBTI, DISC, Jungian, or Keirsey language only for coaching shorthand and translate it back into observable behavior">Communication overlay</option>
                                    <option value="motivation-overlay: use Enneagram or motivation labels as provisional reflection language, not fixed type or hidden motive authority">Motivation overlay</option>
                                    <option value="riasec-fit: use Holland RIASEC only for interest and environment fit guidance, not employment gatekeeping">RIASEC fit</option>
                                </select>
                            </label>
                            <label>
                                <span>Social dynamics lens</span>
                                <select name="social_dynamics_lens">
                                    <option value="none: no formal group-dynamics or clique lens selected">No group lens</option>
                                    <option value="bridge-liaison: protect weak ties, cross-group translation, missing context, and echo-chamber repair">Bridge / liaison</option>
                                    <option value="gatekeeper-representative: audit information flow, access control, external narrative, and resource movement">Gatekeeper / representative</option>
                                    <option value="core-periphery: map center, periphery, access paths, resource flow, and repair routes without ranking human worth">Core-periphery</option>
                                    <option value="clique-risk-audit: audit exclusion, favoritism, conformity pressure, closed-door decisions, resource hoarding, and inclusion repairs">Clique-risk audit</option>
                                    <option value="workplace-clique-review: distinguish vertical, horizontal, mixed, strategic, erratic, apathetic, and conservative workplace clique patterns without legal conclusions">Workplace clique review</option>
                                </select>
                            </label>
                            <label>
                                <span>Research calibration</span>
                                <select name="research_calibration">
                                    <option value="standard-calibration: use evidence boundaries, behavior crosswalks, and ordinary human review before sharing">Standard calibration</option>
                                    <option value="evidence-intensive-calibration: require observable behavior examples, confidence bands, counterexamples, and uncertainty notes before any trait or clique label">Evidence intensive</option>
                                    <option value="human-review-gate: require explicit human review before sharing, importing, or using trait and clique conclusions in team-facing settings">Human review gate</option>
                                    <option value="creative-sandbox-calibration: allow archetype language only when clearly marked as fictional, symbolic, or persona-design material">Creative sandbox</option>
                                </select>
                            </label>
                            <label>
                                <span>Use-case boundary</span>
                                <select name="use_case_boundary">
                                    <option value="persona-design: tune AI persona voice, behavior, and memory without labeling real people">Persona design</option>
                                    <option value="self-reflection: support user-led reflection without diagnosis, prediction, or fixed identity claims">Self-reflection</option>
                                    <option value="team-reflection: review access, pressure, communication, and repair patterns without HR or legal authority">Team reflection</option>
                                    <option value="fictional-character-design: use traits, archetypes, and clique roles only for clearly fictional character work">Fictional character design</option>
                                    <option value="high-stakes-adjacent: produce questions, safeguards, and referral paths instead of conclusions">High-stakes adjacent</option>
                                </select>
                            </label>
                        </div>
                    </section>