II.
StackProfile JSON
Structured · livestack-profile:prompt-engineering-workbench
Prompt Engineering Workbench (TypeScript, React, PostgreSQL, LLM APIs, Redis) json
Inspect the normalized record payload exactly as the atlas UI reads it.
{
"id": "stack-profile:prompt-engineering-workbench",
"_kind": "StackProfile",
"_file": "domain/stack-profiles/deep-stacks-4.yaml",
"_cluster": "domain",
"attributes": {
"displayName": "Prompt Engineering Workbench (TypeScript, React, PostgreSQL, LLM APIs, Redis)",
"description": "A developer-facing prompt engineering workbench that provides a React-based\nplayground for authoring, versioning, and A/B testing prompts against\nmultiple LLM providers. PostgreSQL stores prompt versions, evaluation\ndatasets, and scored results. Redis caches LLM responses for rapid\niteration during prompt development. The TypeScript backend (via Hono)\nproxies LLM API calls with token tracking, cost attribution, and\nlatency measurement. Zod validates prompt templates and evaluation\nschemas. Designed for AI product teams and prompt engineers who need\nstructured experimentation beyond ad-hoc notebook workflows. The\ntradeoff is evaluation subjectivity — automated scoring captures\nsurface-level quality but often requires human rating for nuanced\nprompt quality assessment.\n",
"composes": [
"language:typescript",
"framework:react",
"library:zod",
"library:ioredis",
"framework:hono",
"library:prisma"
]
},
"outgoingEdges": [
{
"from": "stack-profile:prompt-engineering-workbench",
"to": "language:typescript",
"kind": "composed_of"
},
{
"from": "stack-profile:prompt-engineering-workbench",
"to": "framework:react",
"kind": "composed_of"
},
{
"from": "stack-profile:prompt-engineering-workbench",
"to": "library:zod",
"kind": "composed_of"
},
{
"from": "stack-profile:prompt-engineering-workbench",
"to": "library:ioredis",
"kind": "composed_of"
},
{
"from": "stack-profile:prompt-engineering-workbench",
"to": "framework:hono",
"kind": "composed_of"
},
{
"from": "stack-profile:prompt-engineering-workbench",
"to": "library:prisma",
"kind": "composed_of"
},
{
"from": "stack-profile:prompt-engineering-workbench",
"to": "library:zustand",
"kind": "composed_of"
},
{
"from": "stack-profile:prompt-engineering-workbench",
"to": "library:tailwindcss",
"kind": "composed_of"
},
{
"from": "stack-profile:prompt-engineering-workbench",
"to": "role:ml-engineer",
"kind": "used_by_role"
},
{
"from": "stack-profile:prompt-engineering-workbench",
"to": "role:frontend-engineer",
"kind": "used_by_role"
},
{
"from": "stack-profile:prompt-engineering-workbench",
"to": "role:research-engineer",
"kind": "used_by_role"
},
{
"from": "stack-profile:prompt-engineering-workbench",
"to": "workflow:prompt-engineering-iteration",
"kind": "follows_workflow"
},
{
"from": "stack-profile:prompt-engineering-workbench",
"to": "workflow:agent-evaluation-cycle",
"kind": "follows_workflow"
},
{
"from": "stack-profile:prompt-engineering-workbench",
"to": "domain:ml-ai",
"kind": "applies_to"
},
{
"from": "stack-profile:prompt-engineering-workbench",
"to": "domain:software-engineering",
"kind": "applies_to"
},
{
"from": "stack-profile:prompt-engineering-workbench",
"to": "skill-area:prompt-engineering",
"kind": "requires_skill_area"
},
{
"from": "stack-profile:prompt-engineering-workbench",
"to": "skill-area:ai-evaluation",
"kind": "requires_skill_area"
},
{
"from": "stack-profile:prompt-engineering-workbench",
"to": "skill-area:frontend-development",
"kind": "requires_skill_area"
},
{
"from": "stack-profile:prompt-engineering-workbench",
"to": "skill-area:backend-api-design",
"kind": "requires_skill_area"
},
{
"from": "stack-profile:prompt-engineering-workbench",
"to": "skill-area:data-analytics",
"kind": "requires_skill_area"
}
],
"incomingEdges": []
}