-
Notifications
You must be signed in to change notification settings - Fork 0
Modes and Paths
Enreign edited this page Mar 8, 2026
·
3 revisions
The skill uses progressive disclosure — it asks only what it needs.
| Mode | Questions | Best For |
|---|---|---|
| Instant | 0 — infers everything from task description | Quick sizing, "how big is this?", triage |
| Quick | 4 questions + sensible defaults | Fast sizing, backlog grooming, rough estimates |
| Detailed | 13 questions, full control | Sprint commitments, external deadlines, high-stakes |
| Scope | Input | Output |
|---|---|---|
| Single | One task description | Full breakdown for one task |
| Batch | List of tasks (any format) | Summary table + rollup + per-task details |
Combining speed and scope gives six paths:
Instant + Single → Zero questions. Infer, estimate, output.
Instant + Batch → Paste a list, get a table. No questions.
Quick + Single → ~4 questions, immediate output.
Quick + Batch → Paste a list, confirm inferences, get a table.
Detailed + Single → Full questionnaire, rich output.
Detailed + Batch → Full questionnaire for shared defaults, per-task overrides.
- What needs to be done? → Infers complexity and task type
- What type of work? → Task type multiplier (coding, testing, infrastructure, etc.)
- How many humans and agents? → Team scaling
- Automation maturity? → Minutes per agent round
Everything else uses defaults: risk=1.3, integration=15%, review=standard, confidence=80%.
- Risk factors → Risk coefficient (derived from specific concerns)
- Integration complexity → Integration overhead percentage
- Domain familiarity → Domain multiplier
- Review depth → Human review time
- Human fix ratio → How much agent output needs correction
- Confidence level → 50%/80%/90% commitment level
- Definition phase → Cone of uncertainty spread
- Organization context → Org overhead on human time
- Dependencies → Sequencing for batch estimates
The first question is always: "Quick estimate or detailed walkthrough?"
Use Instant when:
- You just need a rough number fast
- You're triaging a backlog and need sizes
- The task description is clear enough to infer type and complexity
Use Quick when:
- You need a rough size for planning
- You're estimating a large backlog
- The tasks are well-understood
Use Detailed when:
- You're committing to a deadline
- The work has significant unknowns
- Stakeholders need confidence-level commitments
- You're estimating a high-risk or high-cost task
Getting Started
Core Concepts
- How It Works
- Task Types
- Agent Effectiveness
- Confidence Levels
- Cone of Uncertainty
- PERT Statistics
- Small Council
Reference
Accuracy
Contributors