skills.rest
highflame-ai
highflame-aiOFFICIAL
Explore all skills by this author →
Website𝕏@javelinaiGitHub

source-driven

Retrieve authoritative documentation and live system state before implementing dependent code.

4|Updated Jul 20, 2026
Search Tags:#architecture#debugging#mcp#code-quality#contract-testing#source-truth
One-click install
npx skills add https://github.com/highflame-ai/ai-factory --skill source-driven-highflame-ai
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: source-driven
Source: https://github.com/highflame-ai/ai-factory/tree/main/skills/source-driven
Command: npx skills add https://github.com/highflame-ai/ai-factory --skill source-driven-highflame-ai

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This skill prevents bugs caused by relying on memory or assumptions when implementing code that depends on external contracts, schemas, or platform invariants.

Core Features & Use Cases

  • Authoritative Querying: Forces the AI to read live platform state, ADRs, or codebase definitions before writing dependent code.
  • Contract Verification: Ensures implementation logic matches actual deployed schemas and service contracts rather than recalled approximations.
  • Use Case: When adding a new database query, use this skill to first describe the table schema via MCP tools to ensure column names and types are accurate, preventing runtime errors.

Quick Start

Use the source-driven skill to verify the billing service API contract by reading the relevant ADR and checking the live schema before implementing the new endpoint.

Frequently Asked Questions about source-driven

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
Why does generated code fail against deployed schemas, and how do I verify live state?▼

Ensure database schema accuracy by querying live table definitions via MCP tools before implementation. This forces the AI to verify exact column names and types, preventing runtime errors.

What is the best way to implement cross-service integration without relying on memory?▼

Yes, you need MCP tools to enforce source-driven code generation. The skill requires integration with MCP tools, ADR directories, and codebase search utilities to retrieve authoritative technical documentation.

How do I prevent AI coding hallucinations when integrating external service contracts?▼

The best way to implement cross-service integration is to read relevant ADRs and check live schemas before writing code. This ensures implementation logic matches actual deployed service contracts.

Do I need MCP tools to enforce source-driven code generation?▼

Generated code fails when it relies on recalled approximations rather than live state. Verify deployed schemas by using MCP tools to query the actual database structure before writing integration logic.

When should I not rely on AI memory for configuration management tasks?▼

Do not rely on AI memory for configuration management when platform invariants or external contracts are involved. Always force retrieval of authoritative documentation to ensure implementation accuracy.

Related Skills

Explore More Software Engineering Skills→
JingyiWngJingyiWng

databricks-docs

Retrieve Databricks documentation pages from the llms.txt index.

Community
Basic
CidLucasCidLucas

supabase-postgres-best-practices

Analyze and optimize Postgres queries and schema designs using Supabase best practices.

Community
Advanced
refractionPOINTrefractionPOINT
29

fleet-payload-tasking

Deploy payloads and shell commands fleet-wide to endpoints with reliable tasking.

Official
Advanced
bordenetbordenet
8

experimental-self-prompting

Generate context-free prompts to reframe stalled analysis.

Community
Advanced