SPINE → Interactive Demo

Task Tier Classifier

Determine the appropriate enforcement tier for any development task based on SPINE's tiered protocol.

Classify Your Task

Tier Reference

TIER 1 Simple

No Enforcement

Direct implementation is acceptable. No subagent or MCP usage required.

Examples: Typo fixes, single-file edits, adding comments
TIER 2 Medium

Recommended

SHOULD use subagents and MCP. Document if skipped.

Examples: Multi-file changes, new features, bug investigation
TIER 3 Complex

Mandatory

MUST use appropriate subagents and MCP servers.

Examples: Architecture decisions, research, UI-heavy work

SPINE's Three Capability Layers

Layer 1: Claude Native Built-in Task tool with subagent_types (Explore, Plan, code-architect, visual-tester)
Layer 2: MCP Servers External tools via MCP protocol (browser-mcp, next-conductor, research-agent)
Layer 3: SPINE Python Custom orchestration (fan_out, pipeline, ToolEnvelope, run_scenario.py)