How to Force-Load a Specific Skill in the PM Skills Marketplace

You can force-load a specific skill in the PM Skills Marketplace by prefixing your prompt with /plugin-name:skill-name or simply /skill-name, which immediately activates the skill without waiting for Claude to detect relevant trigger phrases.

The phuryn/pm-skills repository hosts a collection of PM (Product Management) skills that Claude automatically loads when conversations match specific trigger phrases. When you need immediate access to specialized knowledge—for example, when prioritizing skills over general knowledge—you can use the force-load syntax to bring the skill into context instantly.

Understanding the Force-Load Syntax

According to the [README.md](https://github.com/phuryn/pm-skills/blob/main/README.md) in the phuryn/pm-skills repository, you have two options for force-loading skills:

Full plugin prefix: Use /plugin-name:skill-name to explicitly specify which plugin owns the skill.

Short form: Use /skill-name and Claude will automatically resolve the plugin prefix.

The documentation states: "Skills are loaded automatically when relevant to the conversation — no explicit invocation needed. If needed (e.g., prioritizing skills over general knowledge), you can force loading skills with /plugin-name:skill-name or /skill-name."

The developer guide in [CLAUDE.md](https://github.com/phuryn/pm-skills/blob/main/CLAUDE.md) confirms this: "Skills can be force-loaded with /plugin-name:skill-name or /skill-name."

How Force-Loading Works Internally

Understanding the internal mechanism helps clarify when and why to use force-loading.

Skill Registration

Each skill in the PM Skills Marketplace lives under a directory structure like plugin-name/skills/skill-name/ and contains a SKILL.md file with front matter defining a description and trigger phrases. For example, the resume review skill is located at pm-toolkit/skills/review-resume/SKILL.md, while the strategy red team skill resides at pm-execution/skills/strategy-red-team/SKILL.md.

Auto-Loading vs Force-Loading

Under normal operation, Claude watches user utterances for matches against skill trigger phrases. When a match occurs, the skill is fetched and its knowledge becomes part of the model's context.

Force-loading bypasses this detection phase. By using the / prefix, you tell Claude to load the skill immediately, ensuring its knowledge is available for your next prompt regardless of whether the conversation naturally triggered it.

Practical Force-Loading Examples

Force-Loading with Full Plugin Prefix

When you need to be explicit about which plugin provides the skill, use the full syntax:

User: /pm-toolkit:review-resume
Claude: (skill `review-resume` is now loaded)
User: Here is my resume PDF…

This approach is useful when multiple plugins might have skills with similar names, or when you want to ensure the correct version is loaded from the pm-toolkit plugin.

Using the Short Form Syntax

For quicker access, you can omit the plugin name:

User: /review-resume
Claude: (skill `review-resume` loads automatically)
User: Please tailor this resume for a senior PM role.

Claude automatically resolves the plugin prefix, making this ideal for fast iteration when you know the skill name.

Command Line Usage

When using Codex or Claude Code CLI, first install the plugin containing the skill, then use the force-load syntax in your interactive session:

# Install the plugin that contains the skill

claude plugin install pm-toolkit@pm-skills

# Force-load the skill in an interactive session

claude> /pm-toolkit:review-resume

This ensures the skill is available in your CLI environment before you begin the specific task.

Loading Strategy and Analysis Skills

For complex product management tasks, you might need specialized analysis skills:

User: /pm-execution:strategy-red-team
Claude: (skill `strategy-red-team` is now active)
User: Analyze the assumptions in this product roadmap.

This activates the strategy red team skill from the pm-execution plugin, enabling critical analysis capabilities immediately.

Key Implementation Files

The force-loading mechanism is documented and implemented across several key files in the repository:

• README.md – Contains the "Force-loading skills" section explaining the /plugin-name:skill-name syntax and use cases for prioritizing skills over general knowledge.
• CLAUDE.md – Provides developer-focused documentation stating "Skills can be force-loaded with /plugin-name:skill-name or /skill-name."
• pm-toolkit/skills/review-resume/SKILL.md – Example skill that supports force-loading via /pm-toolkit:review-resume or /review-resume.
• pm-execution/skills/strategy-red-team/SKILL.md – Another example skill illustrating the force-loading mechanism.

Summary

• Force-load syntax: Use /plugin-name:skill-name for explicit loading or /skill-name for automatic plugin resolution.
• Immediate activation: Force-loading bypasses trigger phrase detection, making skill knowledge available instantly.
• File locations: Skills are stored under plugin-name/skills/skill-name/SKILL.md with front matter defining triggers and descriptions.
• Use cases: Ideal when prioritizing specialized PM knowledge over general model knowledge, or when trigger phrases haven't naturally occurred in the conversation.

Frequently Asked Questions

What is the exact syntax for force-loading a skill?

You can force-load a skill using either /plugin-name:skill-name (explicit) or /skill-name (short form). Both commands must appear at the beginning of your prompt. According to the phuryn/pm-skills documentation, Claude automatically adds the plugin prefix when you use the short form.

When should I force-load a skill instead of letting Claude auto-detect it?

Force-loading is recommended when you need to prioritize specific skill knowledge over the model's general knowledge, or when the conversation hasn't naturally triggered the skill's keywords. The README.md specifically mentions this is useful for "prioritizing skills over general knowledge" when you know exactly which PM capability you need.

Can I force-load multiple skills at once?

The documentation does not explicitly mention chaining multiple force-load commands in a single prompt. Based on the syntax structure, you would likely need to load skills sequentially—first activating one skill, then loading the next in subsequent prompts—to ensure each skill's context is properly registered.

Where are skill definitions stored in the repository?

Skill definitions are stored as markdown files under plugin-name/skills/skill-name/SKILL.md. For example, the resume review skill is located at pm-toolkit/skills/review-resume/SKILL.md, and the strategy red team skill is at pm-execution/skills/strategy-red-team/SKILL.md. Each file contains YAML front matter with description and trigger phrases that Claude uses for auto-loading.

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