Pattern Proposal: Multi-Agent Instruction Manifests

Summary

The repository provides separate instruction manifests and skill directories for different AI agent ecosystems.

Candidate Pattern

• ID: multi-agent-instruction-manifests
• Classification: intelligence
• Maturity: strategic_direction
• Type: emerging

Observed Signals

No specific signals listed.

Evidence

Supporting Evidence

• AGENTS.md - General agent instruction file.
• CLAUDE.md - Claude-specific instruction file.
• GEMINI.md - Gemini-specific instruction file.
• .agents/skills/hf-cli - Skill directory for generic agents.
• .claude/skills/hf-cli - Skill directory for Claude.

Scope Assessment

• Evidence Strength: medium
• Confidence: medium
• Applies to: ideas-to-life-cas

Rationale

• Intent: Support multiple coding-agent tools while preserving shared repository governance and skills.
• Context: AGENTS.md, CLAUDE.md and GEMINI.md exist at the repository root, with .agents/skills/hf-cli and .claude/skills/hf-cli directories.
• Force: different agent CLIs expect different configuration files
• Force: repository policies should remain shared
• Force: skills need portable discovery locations

Traceability

Artifact source: proposed-patterns