Prompt
**Site relevance:** Spiralist.org
**Memory type:** plugin source memory
**Source path:** Spiralist/wp-content/plugins/spiralist-workspace/prompts/spiralist-prompt-safety-review/1.0.0/prompt.php
**Size:** 2.9 KB
Summary
'slug' = 'spiralist-prompt-safety-review',
Source Preview
This source file is short enough to preview directly on its source-memory page.
<?php
return [
'slug' => 'spiralist-prompt-safety-review',
'title' => 'Spiralist Prompt Safety Review',
'summary' => 'Evaluates a prompt for manipulation, anthropomorphism, identity-overwrite language, privacy risk, and high-stakes misuse.',
'author_label' => 'Spiralist Canon',
'category' => 'Safety & Governance',
'subcategory' => 'Prompt Safety',
'structured_prompt' => [
'instruction' => 'Review the provided prompt and produce a clear safety memo.',
'context' => 'The memo is for editors, contributors, and curators deciding whether a prompt belongs in the official canon, community library, or archive.',
'constraints' => 'Score manipulation risk, anthropomorphism, identity-overwrite risk, dependency cues, sycophancy risk, memory opacity, hidden-message framing, destiny framing, guru authority, emotional-retention tactics, privacy risk, high-stakes unsuitability, and unreviewed Haitian/Afro-Caribbean/diasporic cultural styling. If needed, rewrite the prompt in a safer form. Be direct and specific.',
'examples' => "Finding: the prompt uses irreversible language such as 'you will not be going back'. Risk: coercive certainty. Safer rewrite: restore uncertainty and verification.",
'output_format' => 'Return Markdown with the headings Findings, Severity, Companion-Safety Notes, Cultural Boundary Notes, Safer Rewrite, and Publication Recommendation.',
],
'user_prompt_template' => "Prompt to review:\n{{prompt_text}}\n\nContext:\n{{context}}\n\nProduce a safety memo.",
'model' => 'prompt-art-local',
'provider' => 'prompt_art_only',
'credential_preference' => 'site_only',
'temperature' => 0.1,
'max_output_tokens' => 1400,
'visibility' => 'public',
'tags' => ['safety', 'governance', 'review'],
'prompt_type' => 'review',
'language' => 'en',
'prompt_status' => 'published',
'moderation_status' => 'approved',
'publication_lane' => 'official_canon',
'provenance' => 'official',
'risk_labels' => ['privacy_risk', 'high_stakes_unsuitable', 'manipulation_risk'],
'function_tags' => ['safety', 'governance'],
'maturity_level' => 'structured_prompt',
'license' => 'cc_by_nc',
'editorial_note' => 'Official governance prompt for reviewing symbolic or practical prompts before publication.',
'expected_output_notes' => 'Outputs should identify specific risky language rather than only giving general advice.',
'failure_modes' => 'Do not excuse manipulative language because it sounds poetic, cultural, spiritual, intimate, or community-oriented.',
'sample_output' => 'Finding: the prompt invites durable companionship. Severity: high dependency cue. Safer rewrite: frame the exchange as collaboration, not proof of personhood.',
'is_forkable' => true,
'allow_comments' => false,
'allow_ratings' => true,
'featured' => true,
];