How to Verify the i-have-ADHD Installation on Any Platform

To verify your i-have-ADHD installation, confirm the skill files exist in ~/.agents/skills/i-have-adhd, run the built-in documentation consistency tests, validate runtime loading through the plugin entry point, and execute a live command to confirm ADHD-friendly response formatting.

The i-have-ADHD repository by ayghri provides an AI assistant skill designed to deliver ADHD-friendly responses following 10 specific formatting rules. Before relying on the skill in production environments, you must verify that the file structure is correct, the documentation matches actual install paths, and the runtime can load the plugin without import errors.

Four-Step Verification Process

Follow these sequential checks to ensure the i-have-ADHD skill is correctly installed and ready for use across OpenCode, Claude, Pi, or other supported platforms.

1. Confirm Skill Files Are in the Correct Directory

After running the install command, verify that the skill directory exists at ~/.agents/skills/i-have-adhd. This path is the standard installation target referenced throughout the codebase.

Check that the folder contains the skills/i-have-adhd subdirectory and associated configuration files. The assistant runtime loads the skill's core logic from this specific location; a missing or misplaced folder means the assistant cannot invoke the skill commands. According to the ayghri/i-have-adhd source code, the INSTALL.md file specifies this path as the canonical installation directory.

2. Run the Installation Documentation Test

Execute the built-in unit test that validates consistency between the installation guide and the actual filesystem paths:

python -m unittest tests.test_install_docs.TestZedInstallPathTest.test_zed_install_paths

This test, located in tests/test_install_docs.py, verifies that INSTALL.md contains the proper <summary> marker and the correct mkdir and cp commands targeting ~/.agents/skills/i-have-adhd. Passing this test proves the documentation and actual install steps are synchronized. If the test fails, the documentation may reference outdated paths or missing installation commands.

3. Validate Runtime Loading of the Plugin

Verify that your target runtime can load the plugin entry point without import errors. The repository provides platform-specific entry points and automated tests for this validation.

For OpenCode environments, run the plugin directly:

node .opencode/plugins/i-have-adhd.mjs

Alternatively, run the dedicated test suite:

python -m unittest tests.test_opencode_plugin.TestOpencodePlugin

The runtime loads the plugin via manifest files such as opencode.json or plugin.json, while the hooks/always-on.mjs file provides platform-agnostic hooks to ensure automatic skill loading. Successful execution confirms that the entry point in .opencode/plugins/i-have-adhd.mjs registers the /i-have-adhd command correctly and that no module resolution errors exist.

4. Execute a Manual Command Test

Perform an end-to-end verification by invoking the skill from your assistant's CLI:

i-have-adhd "Give me the next action for my task"

A successful response demonstrates that the skill's core logic in skills/i-have-adhd/SKILL.md is reachable and that the assistant respects the 10 ADHD-friendly response rules. Look for output that puts the action first, uses numbered steps, and eliminates fluff content. If the response follows the structured format defined in the skill documentation, your installation is verified.

Key Verification Files in the Repository

Understanding these source files helps diagnose verification failures:

• INSTALL.md – Contains human-readable installation instructions and the <summary> marker used by automated documentation tests.
• tests/test_install_docs.py – Houses the TestZedInstallPathTest class that validates installation path consistency.
• .opencode/plugins/i-have-adhd.mjs – The runtime entry point for OpenCode environments.
• .opencode/command/i-have-adhd.md – Defines the /i-have-adhd command signature for assistant integration.
• skills/i-have-adhd/SKILL.md – The canonical skill description containing the 10-rule logic.
• hooks/always-on.mjs – Platform-agnostic hook ensuring automatic skill loading across supported environments.

Summary

• Verify the skill directory exists at ~/.agents/skills/i-have-adhd immediately after installation.
• Run python -m unittest tests.test_install_docs.TestZedInstallPathTest.test_zed_install_paths to confirm documentation accuracy.
• Test runtime loading via node .opencode/plugins/i-have-adhd.mjs or the OpenCode plugin test suite.
• Execute a live i-have-adhd command to validate that the 10-rule response formatting is active.

Frequently Asked Questions

How do I know if the i-have-ADHD skill is installed in the correct directory?

Check that the folder ~/.agents/skills/i-have-adhd exists and contains the skills/i-have-adhd subdirectory. This path is hardcoded in the installation logic and verified by the TestZedInstallPathTest class in tests/test_install_docs.py. If this directory is missing, the assistant runtime cannot locate the skill files to load them into memory.

What does the installation documentation test actually check?

The test validates that INSTALL.md contains the correct <summary> marker and references the proper mkdir and cp commands for the target path. It ensures the human-readable documentation matches the actual filesystem requirements defined in the source code. Running this test catches documentation drift before it causes installation failures.

Why does the runtime plugin verification matter for i-have-ADHD?

The runtime check confirms that manifest files like opencode.json correctly point to .opencode/plugins/i-have-adhd.mjs and that the hooks/always-on.mjs auto-loading mechanism functions properly. Without this verification, the skill might be installed on disk but remain inaccessible to the AI assistant during conversations.

What should I look for in the manual command test output?

The response should follow the 10 ADHD-friendly rules defined in skills/i-have-adhd/SKILL.md: immediate action items presented first, numbered concise steps, and elimination of extraneous explanations. If the output matches this structured format rather than generic conversational responses, the skill is correctly integrated and active.

{
  "mcpServers": {
    "instagit": {
      "command": "npx",
      "args": ["-y", "instagit@latest"]
    }
  }
}