Claude-skill-registry ai-prompt-manager

Expert assistant for managing AI prompts, features, and configuration in the KR92 Bible Voice AI system. Use when creating AI prompts, configuring AI features, managing prompt versions, setting up AI bindings, or working with AI pricing and models. Supports multiple vendors and models for feature flexibility.

install

source · Clone the upstream repo

git clone https://github.com/majiayu000/claude-skill-registry

Claude Code · Install into ~/.claude/skills/

T=$(mktemp -d) && git clone --depth=1 https://github.com/majiayu000/claude-skill-registry "$T" && mkdir -p ~/.claude/skills && cp -r "$T/skills/data/ai-prompt-manager-spectaculous-code-raamattu-nyt" ~/.claude/skills/majiayu000-claude-skill-registry-ai-prompt-manager && rm -rf "$T"

manifest: skills/data/ai-prompt-manager-spectaculous-code-raamattu-nyt/SKILL.md

safety · automated scan (low risk)

This is a pattern-based risk scan, not a security review. Our crawler flagged:

dumps environment variables

Always read a skill's source content before installing. Patterns alone don't mean the skill is malicious — but they warrant attention.

source content

AI Prompt Manager

Quick Start

Core Workflow

Create feature → Define what AI capability is needed
Create prompt template → Write system/user prompts with
```
{{variables}}
```
Create prompt version → Implement the template (allows versioning)
Bind to environment → Connect prompt to dev/stage/prod
Configure provider → Choose vendor and model
Test via Admin panel → Validate response and cost

For SQL patterns, see sql-patterns.md.

Database Schema (Essentials)

Table	Purpose
`ai_features`	Feature registry (key, description)
`ai_prompt_templates`	Prompt structure (task, name)
`ai_prompt_versions`	Prompt variants with `{{variables}}`
`ai_prompt_bindings`	Link prompt version to environment
`ai_feature_bindings`	Link feature to vendor/model/env
`ai_pricing`	Cost per vendor/model
`ai_usage_logs`	Track calls, tokens, cost per user

Full schema: See

Docs/context/db-schema-short.md

Prompt Design

Variable Substitution

Use

{{variable}}

syntax for dynamic content:

system_prompt: "You are a {{role}} assistant for Bible study."
user_prompt_template: "Analyze: {{verse_reference}}"

// At call time
await getPrompt('my_feature', {
  role: 'theological scholar',
  verse_reference: 'John 3:16'
}, 'prod');

Output Schema

Define expected output structure to validate responses:

{
  "type": "object",
  "properties": {
    "summary": {"type": "string"},
    "insights": {
      "type": "array",
      "items": {"type": "string"}
    },
    "references": {
      "type": "array",
      "items": {"type": "string"}
    }
  }
}

Temperature & Tokens

Temperature: 0.0–0.3 (factual), 0.4–0.7 (balanced), 0.8–1.0 (creative)
max_tokens: Set based on expected output length
- Verse lookup: ~50 tokens
- Short analysis: ~200 tokens
- Full commentary: ~1000+ tokens

See providers.md for full guidance.

Vendor & Model Configuration

Vendors:

lovable

openai

anthropic

openrouter

Models are vendor-specific and change over time. Always:

Check current availability in the API docs
Test in dev environment first
Configure pricing in
```
ai_pricing
```
before promoting to prod

See providers.md for selection strategy.

Environment Strategy

dev – Test new prompts, experiment with vendors
stage – Validate cost estimates, pre-production testing
prod – Stable, cost-optimized features

Always follow: dev → stage → prod progression.

Testing AI Features

Go to Admin panel → AI section → Testaus tab
Select feature
Input test variables
Review response, tokens, cost estimate
Iterate on prompt if needed

Monorepo Integration

AI features work across workspace apps:

raamattu-nyt (main Bible app) – Uses most AI features
widgetizer (embed service) – Limited AI features
Edge Functions – Orchestrate calls in
```
ai-orchestrator/
```

All share same

bible_schema

database and configuration.

Skills Handoff

Quota/plan limits → See
```
subscription-system
```
skill
Cost optimization → See
```
performance-auditor
```
skill
Edge Functions → See
```
edge-function-generator
```
skill

References

sql-patterns.md – Common SQL workflows
providers.md – Vendor/model selection and parameters
```
Docs/06-AI-ARCHITECTURE.md
```
– Full system design
```
Docs/context/db-schema-short.md
```
– Database schema details
```
Docs/context/supabase-map.md
```
– Edge Functions & access matrix