⚔️ League API Engineering

Evidence-led, policy-safe engineering guidance for League of Legends data integrations.

"Stop guessing your data architecture. Start engineering it."

league-api-engineering is a product-neutral Agent Skills package built to help AI coders and developers design, audit, debug, research, and govern production integrations with League of Legends data sources.

It acts as an engineering companion that helps an agent choose an appropriate source, record evidence and policy decisions, design explicit adapter boundaries, preserve source provenance, and reject unsafe or incompatible approaches.

(Note: It is not an API client, Riot endorsement, or authorization to access restricted data.)

---

📱 Built with League API Engineering

This skill provides AI agents with the architectural boundaries and provider routing needed to build production-grade League integrations. Below is an example of a Live-Tracking Companion App designed and built entirely through structured planning with this skill:

Live Match Recommendations

Runes & Skill Build Insights

Post-Match Details

🎯 What it covers

• 🔌 Source Selection: Riot Web APIs, Data Dragon, LCU, Live Client Data, and CommunityDragon.
• ⚖️ Provider Governance: Third-party provider due diligence and compatible fallback decisions.
• 🛡️ Compliance & Policy: Source authority, stability, terms, privacy, and attribution distinctions.
• 🏗️ Architecture: Adapter, cache, error, fixture, drift, and canary contracts.
• 🛑 Guardrails: Explicit outcomes for inaccessible, unsupported, or policy-blocked requests.

💡 Usage: Read SKILL.md for routing, then follow its reference links for the selected mode.

---

🚀 Quick Start

Choose one:

• Clone (recommended):

  git clone https://github.com/ZepiGit/league-api-engineering.git
  cd league-api-engineering

  The clone's working directory is the package root — use . as the source in the commands below.

• Release archive: download league-api-engineering-<version>.zip from https://github.com/ZepiGit/league-api-engineering/releases, verify its checksum against the paired .manifest.json's archive_sha256 field, then extract it. Extraction produces a league-api-engineering/ folder — use ./league-api-engineering as the source in the commands below, run from the directory you extracted into.

🛠️ Install in Claude Code or OMP

Claude Code

Copy the package into the local skills directory. Replace SOURCE with . (clone) or ./league-api-engineering (release archive), per the acquisition method above:

Copy-Item -Recurse -Force SOURCE "$HOME\.claude\skills\league-api-engineering"

On POSIX shells:

mkdir -p ~/.claude/skills
cp -R SOURCE ~/.claude/skills/league-api-engineering

Restart or reload the host after installation if it does not discover new skills automatically.

Oh My Pi (OMP)

OMP discovers Claude-compatible local skills from the same ~/.claude/skills/ location. Copy the package there using the platform-appropriate command above, then invoke it by its skill name: league-api-engineering.

✅ Validate

From the package root, run:

python scripts/validate.py .

For a standalone source assessment, adapter contract, or fixture, use each script's --help for its accepted path and invocation. The public validator is designed to fail closed: resolve validation findings before publishing or installing.

📦 Package boundaries

This package has no required dependency on a particular League application or repository. The optional case-study-lolcompanion.md is a bounded example, not an installation requirement.

📄 License

MIT; see LICENSE.