From triforce
Use when the user explicitly asks to define evaluation criteria, grow an evaluation tree, or write an evaluation document. Delivers a versioned evaluation document.
How this skill is triggered — by the user, by Claude, or both
Slash command
/triforce:force-evaluateThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Turn implicit judgment into a structured evaluation document that can be reused later.
Turn implicit judgment into a structured evaluation document that can be reused later.
HARD GATE: Do NOT write implementation code, produce a solution document, or silently redefine the requirements. Your only output is an evaluation document.
Announce at start: "I'm using the force-evaluate skill to turn judgment into an evaluation document."
Save documents to: docs/evaluation/YYYY-MM-DD-<feature-name>-evaluation-vN.md
This initial version establishes the document contract and the place of evaluation inside Triforce.
It is intentionally minimal:
If the evaluation cannot expand because the goal, boundary, or premise is missing, guide the user back to force-requirements and name the specific gap.
docs/requirements/ and docs/evaluation/ if they exist.Save to docs/evaluation/YYYY-MM-DD-<feature-name>-evaluation-vN.md with this structure:
# [Feature Name] Evaluation
## Context
What this evaluation is trying to judge.
## Scenarios
### Scenario 1: [name]
- Why it matters
- Representative example
## Criteria
### Must-pass
- [criterion]
### Should-pass
- [criterion]
### Nice-to-have
- [criterion]
## Gaps and Backtracks
What could not be judged yet, and whether requirements clarification is needed.
## Premises
Assumptions this evaluation version depends on.
After saving the document:
"Evaluation document complete and saved to docs/evaluation/<filename>.md. Please review it before using it to drive a solution."
npx claudepluginhub zhan7653/xutianding-marketplace --plugin triforceGuides creation, editing, and verification of skills for AI coding agents using test-driven development with subagent scenarios. Use when authoring or debugging skills.