基于 SOC 职业分类
用 Codex 或 Claude 帮你安装 复制这段 Prompt,粘贴到 Codex、Claude 或其他助手里,让它检查 Skill 页面并帮你完成安装。
直接命令不会经过审查 Prompt;运行前请先检查来源。
npx skills add https://github.com/benjam3n/reasoningtool --skill dse命令会保持在同一行。复制前请横向滚动并检查完整内容。
想先保存到本地?可下载 SkillsMP 当前能够提供的文件。
正在显示 SKILL.md
Route any input through a branching question tree to narrow down the optimal response strategy before writing. Two stages — PERCEIVE (classify input) then ACT (select response). Covers all prompt types.
Generate exhaustive guesses about user input using ALL search methods with coverage tracking. Guessing is SEARCH through possibility space. Tracks space created vs space covered to ensure comprehensive exploration.
Systematically evaluate and select from a set of guesses, options, or possibilities. Combines ARAW analysis with prioritization to determine which guesses are strong, weak, actionable, or eliminable.
| name | dse |
| description | Evaluate strategies based on the logical soundness of their derivation. |
| output | {"format":"prose"} |
Input: $ARGUMENTS
Evaluate strategies based on the logical soundness of their derivation. Determines whether a strategy is logically valid (conclusions follow), sound (premises true AND inferences valid), complete (no gaps), and necessary vs sufficient.
Strategies that pass deductive evaluation have the highest confidence because their correctness follows from the problem definition itself.
For each inference step in the strategy's derivation:
| Step | Premise(s) | Inference | Conclusion | Valid? |
|---|---|---|---|---|
| 1 | [what's assumed] | [rule applied] | [what follows] | Y/N |
| 2 |
Validity check: Does the conclusion NECESSARILY follow from the premises using the stated inference rule?
For each premise used in the derivation:
| Premise | Type | Truth Status | Evidence |
|---|---|---|---|
| [premise] | axiom/observation/assumption/definition | true/uncertain/false | [source] |
Soundness = validity + true premises. A valid argument with false premises produces unreliable conclusions.
Categorize each weakness found:
| Weakness | Type | Impact | Fixable? |
|---|---|---|---|
| [weakness] | invalid step / uncertain premise / missing step / scope error | strategy collapses / weakened / unaffected | Y/N |
Weakness types:
How strong is the connection between strategy and success?
| Level | Description | Test |
|---|---|---|
| Necessary and sufficient | This is the ONLY strategy that works, AND it guarantees success | No alternative exists AND no failure mode exists |
| Sufficient | This strategy guarantees success, but alternatives exist | Will work, others might too |
| Necessary | Must include this, but it alone isn't enough | Required but not complete |
| Contributory | Increases probability of success but doesn't guarantee | Helps but not decisive |
| Irrelevant | No logical connection to success | Doesn't help |
Based on Steps 1-4:
| Proof Level | Criteria |
|---|---|
| Level 4 — Proven | All steps valid, all premises true, no gaps, necessary and sufficient |
| Level 3 — Strong | All steps valid, most premises true, minor gaps, sufficient |
| Level 2 — Supported | Most steps valid, key premises plausible, some gaps, contributory |
| Level 1 — Suggestive | Logic is reasonable, premises uncertain, significant gaps |
| Level 0 — Unproven | Logic has errors, premises unverified, or major gaps |
For strategies below Level 4, specify what would raise them:
UPGRADE PATH:
Current level: [X]
To reach Level [X+1]:
1. [specific action] — fixes: [weakness] — cost: [effort]
2. [specific action] — fixes: [weakness] — cost: [effort]
Most cost-effective upgrade: [which action gives most level-improvement per effort]
DEDUCTIVE STRATEGY EVALUATION:
Strategy: [what was evaluated]
Validity: [N] of [N] steps valid
Soundness: [N] of [N] premises verified
Completeness: [no gaps / minor gaps / significant gaps]
Necessity level: [necessary and sufficient / sufficient / necessary / contributory]
Proof level: [0-4] — [label]
Weaknesses:
| # | Weakness | Type | Impact | Fix |
|---|----------|------|--------|-----|
| 1 | [weakness] | [type] | [impact] | [fix] |
Upgrade path: [what would raise the proof level]
Overall confidence: [high / moderate / low — with justification]