Cross-Platform Skill Conversion

TL;DR: Converts PMO-APEX skills from Claude Code format (SKILL.md) to equivalent formats for other AI coding assistants: Cursor (.cursorrules), GitHub Codex (AGENTS.md), Google Gemini (system instructions), and others. Preserves skill logic, evidence protocols, and methodology content across platforms.

Principio Rector

El conocimiento de PM no debe estar locked-in a una plataforma. Las skills MOAT contienen expertise de gestión de proyectos que es valiosa independientemente de la herramienta AI. La conversión cross-platform extiende el alcance sin diluir el contenido. El formato cambia; la inteligencia metodológica permanece. [EXPLICIT]

Assumptions & Limits

• Assumes source SKILL.md files are at full MOAT compliance level [PLAN]
• Assumes target platform format specifications are documented and stable [SUPUESTO]
• Breaks when target platform has no equivalent to SKILL.md's allowed-tools constraint
• Not all sections translate 1:1; some platforms lack evidence tagging or quality gate concepts
• Converted skills require validation on target platform — conversion alone does not guarantee functionality
• Licensing constraints may apply to converted skills depending on organizational policy [STAKEHOLDER]

Usage

# Convert skills to Cursor format
/pm:cross-platform-convert $PROJECT --target="cursor" --skills="cost-estimation,budget-baseline"

# Batch convert all skills to Codex format
/pm:cross-platform-convert $PROJECT --target="codex" --skills="all"

# Generate platform compatibility matrix
/pm:cross-platform-convert $PROJECT --type=compatibility-matrix

Parameters:

Parameter	Required	Description
$PROJECT	Yes	Project or framework identifier
--target	Yes	Target platform (cursor, codex, gemini, copilot)
--skills	No	Skills to convert (comma-separated or "all")
--type	No	convert, compatibility-matrix, validate

Service Type Routing

{TIPO_PROYECTO}: All project types can benefit from cross-platform availability. Conversion priority based on target platform usage within the organization.

Before Converting

1. Read the source SKILL.md files to understand content and structure to preserve
2. Read target platform documentation for format specifications and constraints
3. Glob skills/cross-platform-convert/references/*.md for format mapping guides
4. Grep for platform-specific limitations that may affect conversion fidelity

Entrada (Input Requirements)

• Source SKILL.md files to convert
• Target platform(s) and their format specifications
• Platform-specific limitations and capabilities
• Priority skills for conversion
• Target platform tool availability

Proceso (Protocol)

1. Platform analysis — Understand target platform format requirements
2. Skill selection — Choose skills to convert based on priority
3. Format mapping — Map SKILL.md sections to target format structure
4. Content adaptation — Adapt content for target platform constraints
5. Tool mapping — Map allowed-tools to target platform equivalents
6. Conversion execution — Generate converted skill files
7. Validation — Test converted skills on target platform
8. Documentation — Document platform-specific differences
9. Batch processing — Convert remaining skills in batches
10. Maintenance plan — Define sync protocol for skill updates

Edge Cases

1. Target platform has no equivalent capability: Document capability gap. Implement workaround or note limitation in converted skill. Never silently drop functionality. [PLAN]
2. Converted skill too large for target format: Split into multiple files per platform convention. Create index/router file linking sub-skills. [METRIC]
3. Platform format specification changes: Version-pin conversions. Maintain conversion mapping per platform version. Reconvert when specification updates. [SUPUESTO]
4. Licensing prevents conversion: Consult legal/compliance. Document licensing constraints. Convert only permitted skills. [STAKEHOLDER]

Example: Good vs Bad

Good Cross-Platform Conversion:

Attribute	Value
Format mapping	Documented per section per platform
Content preservation	≥90% of skill logic preserved
Tool mapping	Allowed-tools mapped to platform equivalents
Validation	Tested on target platform with sample execution
Documentation	Platform differences documented per skill
Sync protocol	Update procedure for skill changes

Bad Cross-Platform Conversion: Copy-pasting SKILL.md content into target format without format adaptation, tool mapping, or validation. Fails because each platform has different conventions — a verbatim copy may not be parseable, may ignore platform-specific features, and provides no maintenance path. [EXPLICIT]

Validation Gate

• Format mapping documented for every SKILL.md section to target format
• ≥90% of skill logic and content preserved in conversion
• Allowed-tools mapped to target platform equivalents or documented as unavailable
• Converted skills tested on target platform with ≥1 sample execution
• Platform-specific differences documented per converted skill
• No silent functionality drops — every limitation documented
• Batch processing handles ≥10 skills without manual intervention
• Sync protocol defined for propagating source skill updates to conversions
• Skills available on stakeholder-preferred platforms [STAKEHOLDER]
• PM methodology content preserved across all platform conversions [PLAN]

Escalation Triggers

• Target platform lacks equivalent capabilities
• Conversion losing critical skill functionality
• Platform format specification changes
• Licensing constraints on converted skills

Additional Resources

Resource	When to read	Location
Body of Knowledge	Before converting to understand platform ecosystems	references/body-of-knowledge.md
State of the Art	When evaluating new AI platform formats	references/state-of-the-art.md
Knowledge Graph	To understand skill dependencies for conversion order	references/knowledge-graph.mmd
Use Case Prompts	When scoping conversion requirements	prompts/use-case-prompts.md
Metaprompts	To generate format mapping templates	prompts/metaprompts.md
Sample Output	To calibrate expected conversion output	examples/sample-output.md

Output Configuration

• Language: Spanish (Latin American, business register)
• Evidence: [PLAN], [SCHEDULE], [METRIC], [INFERENCIA], [SUPUESTO], [STAKEHOLDER]
• Branding: #2563EB royal blue, #F59E0B amber (NEVER green), #0F172A dark