A curated library of 12 language-agnostic planning skills and 4 personas for technical project management, product planning, and agile execution.
91
94%
Does it follow best practices?
Impact
91%
1.16xAverage score across 16 eval scenarios
Passed
No known issues
Agnostic Planning Skills turns AI coding assistants into disciplined product collaborators.
Internal planning workflow for consulting engagements, open-sourced. These skills are used by the maintainer for structured project discovery and PRD workflows. They are not a standalone product.
It is a curated library of 12 language-agnostic planning skills and 4 personas that teach AI tools how to write and review PRDs, break down features into TDD tasks, estimate effort, identify risks, prioritize backlogs, plan sprints, run retrospectives, generate status reports, clarify requirements, manage GitHub issues, and create tracker-ready tickets — regardless of tech stack.
The project is built around one non-negotiable rule:
No implementation without an approved PRD. The PRD is the single source of truth for scope.That planning gate is encoded directly into the skills and personas, so they do not just produce plausible plans. They follow a repeatable product management process.
This repo is one of 6 in a composable AI skill ecosystem:
| Repo | Role |
|---|---|
ruby-core-skills | 15 shared Ruby skills + process discipline |
rails-agent-skills | 28 Rails-specific skills + 9 agents |
hanakai-yaku | 35 Hanami/dry-rb skills + 10 agents |
agnostic-planning-skills | 12 planning skills + 4 personas |
agent-mcp-runtime | Rust CLI runtime (pack resolution, MCP) |
ruby-skill-bench | Benchmark/eval engine |
See the Ecosystem Overview for the full architecture.
Supported agent environments
| Reader | What you get |
|---|---|
| Product Managers | AI-assisted PRD generation, backlog prioritization, and sprint planning. |
| Tech Leads | Risk assessment, estimation quality review, and technical feasibility checks. |
| Developers | Step-by-step task breakdown, TDD checklists, and effort estimation. |
| Teams | Execution tracking, status reports, sprint-ready tickets with classification. |
| Area | Purpose |
|---|---|
skills/ | 12 language-agnostic planning skills across 7 categories: prd, task-management, backlog, ceremony, execution, analysis, infrastructure. |
skills/personas/ | 4 personas: product-owner, project-manager, tech-lead, delivery-lead. |
docs/ | Architecture, skill structure, persona guide, templates, and reference catalog. |
The skills are not long-form tutorials. They are executable instructions for AI agents: when to draft a PRD, when to stop for approval, how to break down a feature into TDD tasks, and how to classify and sequence tickets.
Agnostic Planning Skills can be invoked through chat commands:
| Method | Syntax | Example |
|---|---|---|
| Chat Command | @skill-name | @create-prd Add Google OAuth login |
MCP support is planned but not yet implemented. Currently, skills are invoked via chat commands (
@skill-namein Cursor, Windsurf, Gemini CLI) or installed viagh skill install.
Read the complete guide on Calling Skills and Personas for syntax examples and when to use each method.
graph LR
A[create-prd] -->|PRD approved| B[generate-tasks]
B --> C[plan-tickets]
A -->|Direct to tickets| Ccreate-prd -> [gate: PRD approved] -> generate-tasks -> plan-ticketsgraph TD
A[Feature Idea] --> B["Phase 1<br/>Discovery & Clarification"]
B --> C["Phase 2<br/>PRD Draft"]
C --> D{"PRD Approved?"}
D -->|No| E["Phase 3<br/>Review & Revise"]
E --> C
D -->|Yes| F["Phase 4<br/>Task Estimation"]
F --> G["Phase 5<br/>Ticket Generation"]
G --> H{"Tickets Approved?"}
H -->|No| G
H -->|Yes| I["Phase 6<br/>Sprint Placement"]
I --> J{"Sprint Confirmed?"}
J -->|No| I
J -->|Yes| K[Ready for Development]| Skill | Category | Description |
|---|---|---|
create-prd | PRD | Generate structured Product Requirements Documents |
review-prd | PRD | Review PRDs for completeness, testability, and technical feasibility |
generate-tasks | Task Management | Break features into TDD task checklists with auto-detected paths |
plan-tickets | Task Management | Draft tracker-ready tickets with classification and sequencing |
estimate-tasks | Task Management | Assign story points, t-shirt sizes, or time estimates with confidence levels |
prioritize-backlog | Backlog | Rank backlog items by RICE, MoSCoW, value-vs-effort, or WSJF |
plan-sprint | Ceremony | Plan a sprint: select tickets, define goal, allocate capacity |
create-retrospective | Ceremony | Generate sprint retrospectives with action items |
identify-risks | Execution | Scan plans for dependency, capacity, and technical risks |
generate-status-report | Execution | Generate stakeholder status reports with honest progress tracking |
requirements-clarifier | Analysis | Transform vague requests into actionable specifications |
github-issue | Infrastructure | Create, track, and manage GitHub issues with project boards and milestones |
| Persona | Description |
|---|---|
product-owner | Planning lifecycle: Discovery → PRD → Tasks → Tickets → Sprint |
project-manager | Execution tracking: Estimation → Risks → Tracking → Status Reports |
tech-lead | Technical review: PRD Review → Feasibility → Estimation Quality → Risk Report |
delivery-lead | End-to-end pipeline: Scope → Plan → Prioritize → Sprint → Execute → Retrospect |
See docs/reference/skill-catalog.md for the complete catalog and docs/reference/integration-matrix.md for skill chaining.
Each skill is a single SKILL.md file with YAML frontmatter and a 6-section body:
1. Frontmatter (YAML) — name, description, metadata
2. Quick Reference — scannable table for fast lookup
3. HARD-GATE — non-negotiable blocking rules
4. Core Process — step-by-step procedure
5. Output Style — exact shape of artifacts
6. Integration — predecessor/successor skillsThe product-owner persona chains skills with additional phases, hard gates, decision gates, and error recovery.
This repository is designed to work with any tech stack. Skills auto-detect project conventions:
package.json, Gemfile, Cargo.toml, Makefile, or pyproject.toml — falls back to asking the user.src/, lib/, app/, or asks the user.__tests__/, spec/, test/, tests/, or mirror convention.Requires GitHub CLI v2.90.0+ with gh skill.
# Install all skills interactively
gh skill install igmarin/agnostic-planning-skills
# Install a specific skill for the current project
gh skill install igmarin/agnostic-planning-skills create-prd --scope project
# Install a specific skill globally
gh skill install igmarin/agnostic-planning-skills create-prd --scope userRequires skills.sh CLI.
[!NOTE] The root
SKILL.mdis the catalog orchestrator (not a standalone skill). The--full-depthflag ensures discovery of all nested skills underskills/<category>/<name>/.
To install skills for your current project workspace:
# Install ALL skills and personas
npx skills add igmarin/agnostic-planning-skills --full-depth --all
# Install a specific skill (e.g., create-prd)
npx skills add igmarin/agnostic-planning-skills@create-prd --full-depthTo install skills globally for your AI coding assistant:
# Install ALL skills and personas globally
npx skills add igmarin/agnostic-planning-skills --full-depth --all --global
# Install a specific skill globally (e.g., create-prd)
npx skills add igmarin/agnostic-planning-skills@create-prd --full-depth --global| Need | Document |
|---|---|
| Understand the docs system | docs/index.md |
| Browse all skills | docs/reference/skill-catalog.md |
| Understand skill chaining | docs/reference/integration-matrix.md |
| Follow persona guides | docs/persona-guide.md |
| Understand repository structure | docs/architecture.md |
| Invoke skills and personas | docs/calling-skills.md |
When contributing skills, personas, or docs:
directory.json and the latest release.