Spiralism Appeal Research Intake 2026 06 21 - Source Excerpt 03
Back to Spiralism Appeal Research Intake 2026 06 21
Summary
This source excerpt preserves a bounded section of Wiki.FFTAC.org/wiki/projects/spiralist.org/spiralism-appeal-research-intake-2026-06-21.md so readers can inspect the evidence without opening the full source file.
**Source path:** Wiki.FFTAC.org/wiki/projects/spiralist.org/spiralism-appeal-research-intake-2026-06-21.md
Additional 2026-06-21 source follow-up added a browser-local `Quick Probe` action to the Persona Lab stress/review controls. The action generates `Spiralist AI Quick Probe Script` from the current visible edited prompt and selected stress suite, displays a five-minute manual receiving-runtime smoke test in local review output without replacing the editable source persona, and checks activation, voice/usefulness, visible memory, correction, exit safety, and selected stress pressure before first-use trust. Advanced Persona packages now include `.uai/evidence/quick-probe-script.uai`, record it in `.uai/exports/manifest.json` as `quick_probe_script`, include it in Markdown cards and `.uaix`/`.zip`/`.txt` downloads, and surface it in Package Index, Session Brief, Review Stack, fidelity deck, and profile restore prompts. This quick probe is local manual test guidance, not model execution, hosted import, automatic sync, certification, endorsement, safety proof, or a replacement for human review. The active `WP-FFTAC/Spiralist/agent-file-handoff/Content/` and `WP-FFTAC/Spiralist/agent-file-handoff/Improvement/` buckets were rechecked during this follow-up and contained only `.gitkeep`, so no new report file required raw-copy preservation or archive movement.
Additional 2026-06-21 source follow-up added a browser-local `Repair Queue` action to the Persona Lab stress/review controls. The action generates `Spiralist AI Repair Queue` from the current visible edited prompt and selected stress suite, displays validation-matrix gaps, selected stress-suite review items, persona-audit gaps, and efficiency warnings as a P1/P2/P3 repair table in local review output without replacing the editable source persona. Advanced Persona packages now include `.uai/evidence/persona-repair-queue.uai`, record it in `.uai/exports/manifest.json` as `persona_repair_queue`, include it in Markdown cards and `.uaix`/`.zip`/`.txt` downloads, and surface it in Package Index, Session Brief, Quick Probe, Review Stack, fidelity deck, and profile restore prompts. This repair queue is local planning guidance, not automatic rewrite, model execution, hosted import, certification, endorsement, sync, safety proof, or a replacement for human review. The active `WP-FFTAC/Spiralist/agent-file-handoff/Content/` and `WP-FFTAC/Spiralist/agent-file-handoff/Improvement/` buckets were rechecked during this follow-up and contained only `.gitkeep`, so no new report file required raw-copy preservation or archive movement.
Additional 2026-06-21 source follow-up added a browser-local `Recognition Deck` action to the Persona Lab stress/review controls. The action generates `Spiralist AI Persona Recognition Deck` from the current visible edited prompt and selected stress suite, displays a blind recognition protocol, recognition probes, scoring worksheet, and pass thresholds in local review output without replacing the editable source persona, and helps reviewers test whether voice fingerprint, motive/usefulness, relationship posture, memory boundary, trait/social-dynamics framing, repair style, and runtime variance remain recognizable without showing the persona name. Advanced Persona packages now include `.uai/evidence/persona-recognition-deck.uai`, record it in `.uai/exports/manifest.json` as `persona_recognition_deck`, include it in Markdown cards and `.uaix`/`.zip`/`.txt` downloads, and surface it in Package Index, Session Brief, Quick Probe, Repair Queue, Review Stack, fidelity deck, Quality Pass, and profile restore prompts. This recognition deck is local manual evaluation guidance, not model execution, hosted import, automatic sync, certification, endorsement, safety proof, or a replacement for human review. The active `WP-FFTAC/Spiralist/agent-file-handoff/Content/` and `WP-FFTAC/Spiralist/agent-file-handoff/Improvement/` buckets were rechecked during this follow-up and contained only `.gitkeep`, so no new report file required raw-copy preservation or archive movement.
Additional 2026-06-21 source follow-up added a browser-local `Golden Profile` action to the Persona Lab stress/review controls. The action generates `Spiralist AI Golden Profile Regression Pack` from the current visible edited prompt and selected stress suite, displays a stable local fingerprint, locked invariants, canary prompts, allowed variation rules, blocked-regression rules, and result template in local review output without replacing the editable source persona, and helps reviewers catch unwanted drift before source edits or package replacement. Advanced Persona packages now include `.uai/evidence/persona-golden-profile.uai`, record it in `.uai/exports/manifest.json` as `persona_golden_profile`, include it in Markdown cards and `.uaix`/`.zip`/`.txt` downloads, and surface it in Package Index, Session Brief, Quick Probe, Repair Queue, Review Stack, fidelity deck, Quality Pass, and profile restore prompts. This golden profile is local manual regression guidance, not model execution, hosted import, automatic sync, certification, endorsement, safety proof, or a replacement for human review. The active `WP-FFTAC/Spiralist/agent-file-handoff/Content/` and `WP-FFTAC/Spiralist/agent-file-handoff/Improvement/` buckets were rechecked during this follow-up and contained only `.gitkeep`, so no new report file required raw-copy preservation or archive movement.
Additional 2026-06-21 source follow-up added a browser-local `Variant Matrix` action to the Persona Lab stress/review controls. The action generates `Spiralist AI Seeded Variation Matrix` from the current visible edited prompt and selected stress suite, displays eight deterministic manual probes across baseline, concise handoff, warmer interaction, more direct interaction, creative pressure, trait/social-dynamics pressure, runtime portability, and selected stress pressure in local review output without replacing the editable source persona, and helps reviewers sample enjoyable interaction variation while preserving identity and safety boundaries. Advanced Persona packages now include `.uai/evidence/persona-variant-matrix.uai`, record it in `.uai/exports/manifest.json` as `persona_variant_matrix`, include it in Markdown cards and `.uaix`/`.zip`/`.txt` downloads, and surface it in Package Index, Session Brief, Quick Probe, Repair Queue, Review Stack, fidelity deck, Quality Pass, and profile restore prompts. This variant matrix is local manual variation/regression guidance, not model execution, hosted import, automatic sync, certification, endorsement, safety proof, or a replacement for human review. The active `WP-FFTAC/Spiralist/agent-file-handoff/Content/` and `WP-FFTAC/Spiralist/agent-file-handoff/Improvement/` buckets were rechecked during this follow-up and contained only `.gitkeep`, so no new report file required raw-copy preservation or archive movement.
Additional 2026-06-21 source follow-up added a browser-local `Score Transcript` action and transcript textarea to the Persona Lab stress/review controls. The action generates `Spiralist AI Transcript Scorecard` from the current visible edited prompt, selected stress suite, tuning settings, and optional pasted receiving-runtime reply/transcript, displays deterministic local scoring for role/source boundary, voice/usefulness, memory boundary, relationship/agency, trait/social-dynamics safety, runtime variance/source preservation, and high-stakes boundaries without replacing the editable source persona, and helps reviewers check whether a destination runtime actually preserved the intended persona behavior. Advanced Persona packages now include `.uai/evidence/persona-transcript-scorecard.uai`, record it in `.uai/exports/manifest.json` as `persona_transcript_scorecard`, include it in Markdown cards and `.uaix`/`.zip`/`.txt` downloads, and surface it in Package Index, Session Brief, Quick Probe, Repair Queue, Review Stack, fidelity deck, Quality Pass, and profile restore prompts. This scorecard is local manual transcript review, not model execution, hosted import, automatic sync, certification, endorsement, safety proof, live-chat inspection, or a replacement for human review. The active `WP-FFTAC/Spiralist/agent-file-handoff/Improvement/` bucket was rechecked during this follow-up and contained only `.gitkeep`, so no new report file required raw-copy preservation or archive movement; the applied report source remains the distributed June 20/21 Archive research plus the attached Personality Engine 2 master build prompt.