How to Brainstorm and Structure Team-Level OKRs Using the PM Skills Marketplace

The pm-skills repository provides a reproducible, AI-native workflow for generating team-level OKRs through the pm-execution plugin, which combines the brainstorm-okrs skill for domain knowledge with the /plan-okrs command for interactive execution.

The pm-skills repository by phuryn is a collection of AI-powered plugins designed to standardize product management workflows. To effectively brainstorm and structure team-level OKRs, the repository offers a dedicated execution plugin that transforms abstract goal-setting theory into structured, version-controlled markdown documents.

Architecture of the OKR Workflow System

The OKR workflow in pm-skills follows a modular architecture separating domain knowledge from execution logic. This design ensures that the conceptual framework for writing Objectives and Key Results remains portable across AI platforms while the interactive workflow adapts to specific contexts.

The Skill Definition (pm-execution/skills/brainstorm-okrs/SKILL.md)

The brainstorm-okrs skill serves as the knowledge base for the entire workflow. Defined in pm-execution/skills/brainstorm-okrs/SKILL.md, this file contains the conceptual framework for writing effective OKRs, including the relationship between Objectives, Key Results, KPIs, and North Star Metrics. The skill provides a step-by-step guide for generating three alternative OKR sets and includes links to further reading on goal-setting best practices.

The Command Interface (pm-execution/commands/plan-okrs.md)

The plan-okrs command wraps the skill in an interactive CLI-style flow. Located at pm-execution/commands/plan-okrs.md, this file defines the invocation syntax, workflow steps, and validation checklist. When executed, the command gathers context about your team and company goals, generates three alternative OKR sets, validates them against the skill's criteria, and returns a ready-to-publish markdown file.

How to Generate Team-Level OKRs

Implementing the OKR workflow requires installing the plugin from the pm-skills marketplace and invoking the planning command with your specific context.

Install the pm-execution Plugin

First, add the marketplace to your AI assistant environment and install the execution plugin that contains the OKR workflow:

# Add the marketplace

claude plugin marketplace add phuryn/pm-skills

# Install the execution plugin (includes the OKR workflow)

claude plugin install pm-execution@pm-skills

Execute the Planning Command

Once installed, invoke the OKR planner by providing your team name, time period, and company-level objective:

/plan-okrs Growth team Q2 — company goal is 50% ARR increase

The command processes this input through the brainstorm-okrs skill, generating structured output that aligns your team objectives with the broader company goal.

Review the Structured Output

The command produces a markdown file named OKRs-[team]-[quarter].md containing a standardized format for team-level OKRs. Each output includes:

• Concise Objectives that align with company strategy
• Three measurable Key Results per Objective with baseline and target values
• Rationale and alignment mapping showing the relationship between company Objectives, Team Objectives, and specific Key Results
• Owner assignments for accountability

Anatomy of the OKR Output Format

The pm-skills framework enforces a specific structure that ensures every OKR is measurable and actionable. Here is an example of the generated output format:

## Team OKRs: Growth team — Q2

**Aligned to**: Increase ARR by 50% (Company Objective)

### Objective 1: Accelerate customer acquisition in high-value segments

| # | Key Result                              | Baseline | Target | Owner |

|---|----------------------------------------|----------|--------|-------|
| 1 | New paid accounts from target segment  | 120      | 260    | Alice |
| 2 | Conversion rate (trial → paid)         | 15%      | 24%    | Bob   |
| 3 | CAC (cost per acquisition) ↓ 20%       | $120     | $96    | Carla |

### Objective 2: Boost product-led growth loops

| # | Key Result                              | Baseline | Target | Owner |

|---|----------------------------------------|----------|--------|-------|
| 1 | Monthly active users (MAU) ↑ 30%       | 18k      | 23k    | Dan   |
| 2 | Referral-driven sign-ups ↑ 40%          | 500      | 700    | Eve   |
| 3 | Activation rate (first-week) ≥ 75%     | 62%      | 75%    | Frank |

### Objective 3: Strengthen revenue predictability

| # | Key Result                              | Baseline | Target | Owner |

|---|----------------------------------------|----------|--------|-------|
| 1 | Net revenue retention (NRR) ≥ 115%    | 108%     | 115%   | Grace |
| 2 | Upsell revenue from existing customers | $80k     | $120k  | Hank  |
| 3 | Churn rate ↓ 15%                        | 8%       | 6.8%   | Ivy   |

Iterative Refinement and Follow-Up Actions

After generating the initial OKR set, the /plan-okrs command offers three follow-up actions to refine the output:

• Adjust ambition levels – Tune the difficulty of Key Results based on historical performance data
• Create a metrics dashboard – Generate tracking mechanisms for monitoring progress toward each Key Result
• Draft a stakeholder update – Prepare communication materials to align team members and leadership on the new OKRs

These prompts enable rapid iteration without leaving the AI environment, ensuring the final OKR document reflects realistic yet ambitious goals.

Summary

• The pm-execution plugin in the pm-skills repository provides a complete workflow to brainstorm and structure team-level OKRs through the combination of the brainstorm-okrs skill and /plan-okrs command.
• The skill file at pm-execution/skills/brainstorm-okrs/SKILL.md contains the domain knowledge and step-by-step instructions for generating valid OKR sets.
• The command file at pm-execution/commands/plan-okrs.md orchestrates the interactive workflow, producing markdown outputs with baseline metrics, targets, and owner assignments.
• Installation requires adding the phuryn/pm-skills marketplace and installing the pm-execution plugin via claude plugin install pm-execution@pm-skills.
• The system generates three alternative OKR sets per execution and includes follow-up prompts for adjusting ambition levels, creating dashboards, and drafting stakeholder communications.

Frequently Asked Questions

How does the pm-skills repository ensure OKRs are aligned with company objectives?

The brainstorm-okrs skill defined in pm-execution/skills/brainstorm-okrs/SKILL.md explicitly requires mapping each team Objective to a specific company Objective before generating Key Results. The output template includes an "Aligned to" field that documents this relationship, ensuring vertical alignment throughout the organization.

Can I customize the number of Key Results per Objective in the output?

The /plan-okrs command generates three Key Results per Objective by default, following standard OKR best practices as defined in the skill file. However, because the skill is a standalone markdown file, you can modify pm-execution/skills/brainstorm-okrs/SKILL.md to adjust the generation instructions before invoking the command.

What AI platforms support the pm-skills OKR workflow?

Because the brainstorm-okrs skill is a standalone markdown file, it can be loaded by any AI assistant that reads the repository. The /plan-okrs command is specifically designed for Claude Code, Claude Cowork, and the Codex CLI, making the workflow portable across these environments while maintaining version control via the GitHub repository.

How does the validation step in the OKR generation process work?

The plan-okrs command at pm-execution/commands/plan-okrs.md includes a validation checklist that verifies each Key Result is measurable, has an assigned owner, and includes specific baseline and target values. This validation occurs after generation but before presenting the final markdown output, ensuring all OKRs meet the criteria defined in the skill file.

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