301dc13d0e
Neurometric markets a task-specialized PR-Summarizer SLM (Qwen3-4B +
specialist system prompt) on a free tier. Adding it as a 4th voice in
the existing A/B reviewer panel to see whether a specialized small
model catches things the 3 general-purpose voices miss.
The SLM does not honor OpenAI tool-calling, so it cannot return our
structured submit_review schema. Wired as a "prose voice":
- MODELS tuple gains a 6th provider_type field
("openrouter_tools" | "neurometric_prose")
- call_model() dispatches; new call_neurometric_prose() path hits
api.neurometric.ai/v1 without tools, embeds raw markdown output,
and infers verdict from keyword heuristic
- Render branch in PR comment marks the verdict as heuristic and
notes tokens/cost are not tracked (free tier)
- Never the primary voice; heading still driven by Haiku
Also hardens the shared render path against models that omit the
optional 'summary' field (rev.get('summary') fallback).
Requires new repo secret NEUROMETRIC_API_KEY (already set on all four
peptiderepo repos). Voice self-skips with inline error if missing.
Agent-Session: cowork-neurometric-qa-voice-20260415
Co-authored-by: peptiderepo <peptiderepo@users.noreply.github.com>
=== Peptide Search AI === Contributors: terence Tags: peptide, database, search, AI, science, research, openrouter Requires at least: 5.8 Tested up to: 6.7 Requires PHP: 7.4 Stable tag: 4.1.0 License: GPLv2 or later Searchable peptide database with AI-powered auto-population. Visitors search for peptides — if one doesn't exist, it's automatically validated and created with comprehensive AI-generated scientific content in the background. == Description == Peptide Search AI turns your WordPress site into an intelligent, self-growing peptide database. **How it works:** 1. Visitors search for a peptide using the search form. 2. If found, results appear instantly with molecular data. 3. If not found, the plugin automatically validates the name using a lightweight AI call, creates a placeholder, and queues comprehensive content generation in the background via WP-Cron. 4. The visitor sees a "currently being added" message and can check back shortly. 5. Generated entries include: amino acid sequence, molecular weight/formula, mechanism of action, research applications, safety profile, dosage info, and references. 6. PubChem cross-referencing validates molecular data when available. **Features:** * AJAX search with instant results * Automatic peptide name validation (lightweight AI check before full generation) * Background content generation via WP-Cron (non-blocking) * Custom Post Type with structured scientific fields * AI content generation via OpenRouter — access hundreds of models (Gemini, Llama, DeepSeek, Mistral, Qwen, and more) * Separate model configuration for validation (cheap/fast) and generation (full) * PubChem PUG REST API integration for verified molecular data * Rate limiting (10 requests/IP/hour) to prevent abuse * Transient-based caching for validation and PubChem lookups * Admin settings for API key, models, auto-publish behavior, PubChem toggle * REST API endpoint for headless/external access (read-only) * Conditional asset loading (only on pages with the shortcode or peptide singles) * Source badges (AI-Generated / PubChem Verified / Manually Curated) * Mobile-responsive design == Installation == 1. Upload the `peptide-search-ai` folder to `/wp-content/plugins/` 2. Activate the plugin through the Plugins menu 3. Go to Settings > Peptide Search AI 4. Enter your OpenRouter API key (get one at openrouter.ai/settings/keys) 5. Optionally set separate models for validation and generation 6. Configure auto-publish and PubChem settings 7. Add `[peptide_search]` to any page or post == Shortcode == `[peptide_search]` — Displays the peptide search form. Optional attributes: * `placeholder` — Custom placeholder text for the search input. Example: `[peptide_search placeholder="Search our peptide database..."]` == REST API == Search endpoint: `GET /wp-json/peptides/v1/search?q=BPC-157` Returns JSON with matching published peptides, their molecular data, and permalink URLs. This endpoint is read-only and does not trigger auto-generation.