MCP Prompts

ido4 provides 8 MCP prompts — portable intelligence frameworks that work with any MCP-compatible AI client, not just Claude Code. Each prompt generates methodology-specific reasoning using the active profile’s terminology and principles.

Prompts vs Skills

Aspect	Skills	Prompts
Platform	Claude Code only	Any MCP client
Invocation	/ido4dev:standup	MCP prompt protocol
Features	Memory, file access, hooks	Tool calls only
Count	21	8

If you’re using Claude Code with the plugin, use skills. If you’re using another MCP client (Cursor, Windsurf, etc.), use prompts.

Available Prompts

standup

Governance-aware briefing. Analyzes container phase, detects blockers with cascade reasoning, identifies temporal patterns from the audit trail, and recommends the highest-impact action.

Arguments: None

plan-{container}

Container composition with principle-aware constraints. The prompt name adapts to your methodology: plan-wave (Hydro), plan-sprint (Scrum), plan-cycle (Shape Up). Framework for dependency analysis, capacity estimation using real throughput data, and validation against the active profile’s principles.

Arguments: containerName (optional)

board

Flow intelligence report. Detects blocked cascades, false statuses, review bottlenecks, container fragmentation, cycle time outliers, and agent coordination issues.

Arguments: containerName (optional)

compliance

Three-part compliance assessment. Quantitative scoring (0-100 with category breakdown), structural audit of governance principles (varies by methodology), and cross-referenced synthesis with actor patterns and temporal trends.

Arguments: None

health

Quick governance dashboard. RED/YELLOW/GREEN across three dimensions (flow, governance, team) with compact metrics and suggested next action.

Arguments: None

retro

Container retrospective. Analyzes real throughput, cycle time trends, measured blocking time, actor patterns, and governance quality. Every recommendation is data-backed. Adapts to methodology: wave retro (Hydro), sprint retro (Scrum), cycle retro (Shape Up).

Arguments: containerName (optional)

review

Container review for stakeholder communication. Deliverable assessment, outcome vs plan analysis, quality metrics, and forward-looking analysis.

Arguments: containerName (optional)

execute-task

Specs-driven task execution guidance. 8-phase framework: specification comprehension, upstream context interpretation, downstream awareness, pattern detection, work execution (methodology-specific principles), escalation protocol, context capture, and completion verification.

Arguments: issueNumber (required)

Methodology-Aware Generation

Prompts aren’t just localized — they use fundamentally different reasoning frameworks per methodology:

• Hydro prompts reason about wave phases, epic integrity, and dependency coherence
• Scrum prompts reason about sprint goals, burndown trajectory, and DoR/DoD per work item type
• Shape Up prompts reason about hill chart positions, appetite consumption, scope hammering, and circuit breaker countdown

The prompt generators read the active methodology profile and produce methodology-native content.

Using Prompts

In any MCP client that supports prompts:

// Request the standup prompt
GET /prompts/standup

// Request plan-wave with arguments
GET /prompts/plan-wave?containerName=wave-003-advanced

The prompt returns a structured reasoning framework that the AI client uses to compose tool calls and analyze results.