Size project tasks with AI
Estimate effort and complexity separately, making scope, success criteria, assumptions and dependencies explicit before sizing the work.
Size project tasks with AI by separating the amount of work from how difficult or uncertain it is. The estimatrix skill separates a relative T-shirt size from a low, medium or high complexity judgement.
It clarifies scope, success criteria, constraints and dependencies before sizing. Distinct interpretations or adjacent cleanup remain separate, and very large work becomes a decomposition prompt. The result is useful for comparison and planning discussions, not a reliable promise of hours or delivery dates. Calibrate it against completed work and the team’s actual context rather than treating a model’s first label as an objective measurement.
What it does
- Clarify scope and completion criteria.
- Separate size from complexity.
- Compare alternative interpretations.
- Keep adjacent work in separate rows.
- Decompose oversized tasks and dependencies.
Install and use
- Install the folder in a SKILL.md-capable host. No separate account is needed. Supply task descriptions, constraints and a comparable completed task, omitting unnecessary private information.
- Ask: “Size these tasks with separate effort and complexity, asking about material ambiguities first.” State how completion will be verified.
- Review assumptions and confidence. Request a rough result explicitly if appropriate, then compare the labels with actual completed work for future calibration.
FAQ
Source
Made by Information Logistics. Original skill: estimatrix, published on GitHub under the MIT licence.
View on GitHub