agentleFS
trysonartrysonar/skills/AGENTS.md

skills

AGENTS.md0 starsChanged 42 days agoMITView on GitHub

This repo contains App Store Optimization skills following the Agent Skills open standard, backed by the Sonar API and MCP server for live data. It is also a Claude plugin (sonar-aso) and a one-plugin marketplace (trysonar). These are the contributor instructions for any coding agent working in this repo, Claude Code included. A root CLAUDE.md is deliberately absent: Claude plugins don't load it, and claude plugin validate warns about it. Compatible directories: .claude/skills/, .cursor/skills/, .agents/skills/, .codex/skills/ The description drives discovery…

The file

Files
AGENTS.md

Agent guidelines

This repo contains App Store Optimization skills following the Agent Skills open standard, backed by the Sonar API and MCP server for live data. It is also a Claude plugin (sonar-aso) and a one-plugin marketplace (trysonar).

These are the contributor instructions for any coding agent working in this repo, Claude Code included. A root CLAUDE.md is deliberately absent: Claude plugins don't load it, and claude plugin validate warns about it.

Structure

SKILL.md                         — root router (for agents importing the repo root)
skills/<name>/SKILL.md           — skill instructions (required, <500 lines)
tools/REGISTRY.md                — capability matrix: which skill uses which tool
.mcp.json                        — hosted Sonar MCP server bundled with the plugin
.claude-plugin/plugin.json       — Claude plugin manifest (name, version, metadata)
.claude-plugin/marketplace.json  — marketplace catalog, so `/plugin marketplace add trysonar/skills` works

Compatible directories: .claude/skills/, .cursor/skills/, .agents/skills/, .codex/skills/

SKILL.md format

---
name: skill-name          # Must match directory, lowercase+hyphens, 1-64 chars
description: ...          # Trigger phrases + scope boundaries, 1-1024 chars
metadata:
  version: 1.1.0
---

The description drives discovery — an agent reads all descriptions to decide which skill to load. Include "When the user wants to...", trigger keywords, and "For X, see Y" boundaries.

Writing style

  • Direct, actionable, second person
  • Sentence case headings, no emoji
  • Tables for comparisons, numbered lists for steps
  • Real tool names and real parameters only — never invent Sonar fields or endpoints
  • Plans are "Indie", "Agency", and "API Credits" — never "Full plan" or "Agent plan"
  • State what a tool requires (plan, sign-in, credits) plainly; don't pitch upgrades or purchases inside skills

Credentials

Nothing in this repo needs, reads, or stores an API key. Skills never expand key environment variables in examples or commands, and never ask the user to paste a key into the conversation. In Claude, the bundled MCP server (.mcp.json, no headers) serves the keyless free tier and authenticates everything else through Sonar's OAuth sign-in. Keep it that way: don't add a key header or a key userConfig option — the free tier only applies when no key is sent, and an invalid key returns 401.

Ground truth

The Sonar API and MCP server are the source of truth. Verify tool behavior against:

The hosted server at https://trysonar.app/mcp exposes every tool except sonar_export_screenshots, which writes to the local filesystem and only exists in the stdio package.

Maintenance rule

The skills describe the live API. When a v1 endpoint or MCP tool changes, this repo is part of the update checklist: update the affected skills and tools/REGISTRY.md, bump version in .claude-plugin/plugin.json (and the touched skills' metadata.version), and run claude plugin validate . — it must pass with zero warnings.

App store reference

iOS limits: title 30, subtitle 30, keyword field 100 (comma-separated, no spaces), description 4000 (not indexed), promotional text 170.

Android limits: title 30, short description 80, description 4000 (all indexed).

Key rules: title carries the highest keyword weight on both stores. On iOS never repeat a keyword across indexed fields and use singular forms. On Android repeat core terms 2–3x naturally in the description.

Commits

feat(skill-name): ... / fix(skill-name): ... / docs: ...

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.