lifeskills

内容来源：README.md（说明文档） · 原始地址 · 查看安装指南

原始内容

# Lifeskills

A curated collection of non-coding skills for AI agents focused on business-critical communication, strategy, negotiation, and influence.

These skills follow a shared quality contract for decision-ready outputs and explicit handling of uncertainty.

Language of skill content: English.

## Skills included

## Negotiation and influence

### [negotiation-voss-tactical-empathy](https://github.com/santos-sanz/lifeskills/tree/HEAD/skills/negotiation-voss-tactical-empathy)
Negotiation planning and execution with tactical empathy, strategy/script separation, BATNA/ZOPA framing, and conditional concessions.

### [persuasion-cialdini-influence-design](https://github.com/santos-sanz/lifeskills/tree/HEAD/skills/persuasion-cialdini-influence-design)
Ethical persuasion design using Cialdini principles with principle-to-evidence-to-claim traceability.

## Structured communication

### Executive Communication

#### [pyramid-principle-structured-communication](https://github.com/santos-sanz/lifeskills/tree/HEAD/skills/pyramid-principle-structured-communication)
Answer-first executive storytelling for board updates, recommendations, and decision memos.

### Mindset & Personal Strategy

#### [luffy-mindset](https://github.com/santos-sanz/lifeskills/tree/HEAD/skills/luffy-mindset)
Applies an audacious, practical "Luffy Mindset" for bold execution plans when explicitly requested.
*   **Best for:** Breaking hesitation, pursuing ambitious goals, and converting fear into action.
*   **Key Capabilities:** Fearless reframing, 3-7 step action planning, risk countermoves, and 24-hour commitment starts.

### Leadership Execution

#### [high-agency](https://github.com/santos-sanz/lifeskills/tree/HEAD/skills/high-agency)
Applies George Mack's High Agency framework for founder and leadership execution under uncertainty.
*   **Best for:** High-stakes decisions, execution bottlenecks, ambiguous constraints, and cross-functional deadlocks.
*   **Key Capabilities:** Converting vague blockers into specific constraints, generating high-agency options, and driving 24-72h action plans.

## Consulting frameworks

### [consulting-issue-tree-mece](https://github.com/santos-sanz/lifeskills/tree/HEAD/skills/consulting-issue-tree-mece)
MECE issue-tree decomposition with formal validation gates and prioritized analysis backlog mapping.

### [consulting-hypothesis-driven-80-20](https://github.com/santos-sanz/lifeskills/tree/HEAD/skills/consulting-hypothesis-driven-80-20)
Hypothesis-driven planning with falsifiable design, kill criteria, and 80/20 prioritization.

### [consulting-market-competition-analysis](https://github.com/santos-sanz/lifeskills/tree/HEAD/skills/consulting-market-competition-analysis)
Market sizing and competitive analysis with top-down/bottom-up triangulation and uncertainty-aware scenarios.

### [consulting-portfolio-growth-strategy](https://github.com/santos-sanz/lifeskills/tree/HEAD/skills/consulting-portfolio-growth-strategy)
Portfolio allocation and growth sequencing using BCG, GE/McKinsey, and Ansoff logic.

## Shared output contract

Most decision-oriented skills use the same section order:
1. Context
2. Decision or Recommendation
3. Analysis
4. Risks
5. Next Actions
6. Assumptions

Specialized execution skills such as `high-agency` and `luffy-mindset` use their canonical templates instead of the generic contract.

This improves consistency across skills and makes outputs easier to review in high-stakes workflows.

## Quality system

The repository includes a reusable evaluation framework in [`quality/`](https://github.com/santos-sanz/lifeskills/tree/HEAD/quality):
- `quality/rubric.md`: common scoring rubric (1-5).
- `quality/test-prompts.md`: 4 test prompts per skill (36 total).
- `quality/eval-log-template.md`: logging template for manual evaluations.
- `scripts/lint_skills.py`: local lint gate for skill structure, catalog coverage, and prompt coverage.

Target acceptance thresholds:
- Per-skill average >= 4.2
- No criterion below 3.5
- Zero critical guardrail failures

## Validation workflow

Run `python scripts/lint_skills.py` after any skill edit. Fix every reported issue before staging, committing, or pushing.

## Installation

```bash
npx skills add https://github.com/santos-sanz/lifeskills
```

Once installed, these skills are automatically available to the agent. You can invoke them explicitly or rely on the agent's intent detection.

**Examples:**

*   *"Help me negotiate a vendor renewal for 2026 using tactical empathy."*
*   *"Draft a persuasive email to the CFO using Cialdini's principles."*
*   *"Create a MECE issue tree to analyze why user retention is dropping."*
*   *"We are blocked on launch and everyone says it's impossible. Build a high-agency decision memo for the next 72 hours."*

## Repository structure

```text
skills/
  <skill-name>/
    SKILL.md
    references/
    templates/
    examples/
scripts/
  lint_skills.py
quality/
  rubric.md
  test-prompts.md
  eval-log-template.md
```

## Design principles used in this version

This iteration integrated patterns commonly used in high-quality agent skills ecosystems:
- Prompt contracts for predictable outputs.
- Rubric-driven evaluation for quality control.
- Explicit uncertainty and assumption tagging.
- Ethical guardrails for persuasion and negotiation tasks.
- Local lint gating before publish.

Lifeskills

A curated collection of non-coding skills for AI agents focused on business-critical communication, strategy, negotiation, and influence.

These skills follow a shared quality contract for decision-ready outputs and explicit handling of uncertainty.

Language of skill content: English.

Skills included

Negotiation and influence

negotiation-voss-tactical-empathy

Negotiation planning and execution with tactical empathy, strategy/script separation, BATNA/ZOPA framing, and conditional concessions.

persuasion-cialdini-influence-design

Ethical persuasion design using Cialdini principles with principle-to-evidence-to-claim traceability.

Structured communication

Executive Communication

pyramid-principle-structured-communication

Answer-first executive storytelling for board updates, recommendations, and decision memos.

Mindset & Personal Strategy

luffy-mindset

Applies an audacious, practical "Luffy Mindset" for bold execution plans when explicitly requested.

Best for: Breaking hesitation, pursuing ambitious goals, and converting fear into action.
Key Capabilities: Fearless reframing, 3-7 step action planning, risk countermoves, and 24-hour commitment starts.

Leadership Execution

high-agency

Applies George Mack's High Agency framework for founder and leadership execution under uncertainty.

Best for: High-stakes decisions, execution bottlenecks, ambiguous constraints, and cross-functional deadlocks.
Key Capabilities: Converting vague blockers into specific constraints, generating high-agency options, and driving 24-72h action plans.

Consulting frameworks

consulting-issue-tree-mece

MECE issue-tree decomposition with formal validation gates and prioritized analysis backlog mapping.

consulting-hypothesis-driven-80-20

Hypothesis-driven planning with falsifiable design, kill criteria, and 80/20 prioritization.

consulting-market-competition-analysis

Market sizing and competitive analysis with top-down/bottom-up triangulation and uncertainty-aware scenarios.

consulting-portfolio-growth-strategy

Portfolio allocation and growth sequencing using BCG, GE/McKinsey, and Ansoff logic.

Shared output contract

Most decision-oriented skills use the same section order:

Context
Decision or Recommendation
Analysis
Risks
Next Actions
Assumptions

Specialized execution skills such as high-agency and luffy-mindset use their canonical templates instead of the generic contract.

This improves consistency across skills and makes outputs easier to review in high-stakes workflows.

Quality system

The repository includes a reusable evaluation framework in quality/:

quality/rubric.md: common scoring rubric (1-5).
quality/test-prompts.md: 4 test prompts per skill (36 total).
quality/eval-log-template.md: logging template for manual evaluations.
scripts/lint_skills.py: local lint gate for skill structure, catalog coverage, and prompt coverage.

Target acceptance thresholds:

Per-skill average >= 4.2
No criterion below 3.5
Zero critical guardrail failures

Validation workflow

Run python scripts/lint_skills.py after any skill edit. Fix every reported issue before staging, committing, or pushing.

Installation

npx skills add https://github.com/santos-sanz/lifeskills

Once installed, these skills are automatically available to the agent. You can invoke them explicitly or rely on the agent's intent detection.

Examples:

"Help me negotiate a vendor renewal for 2026 using tactical empathy."
"Draft a persuasive email to the CFO using Cialdini's principles."
"Create a MECE issue tree to analyze why user retention is dropping."
"We are blocked on launch and everyone says it's impossible. Build a high-agency decision memo for the next 72 hours."

Repository structure

skills/
  <skill-name>/
    SKILL.md
    references/
    templates/
    examples/
scripts/
  lint_skills.py
quality/
  rubric.md
  test-prompts.md
  eval-log-template.md

Design principles used in this version

This iteration integrated patterns commonly used in high-quality agent skills ecosystems:

Prompt contracts for predictable outputs.
Rubric-driven evaluation for quality control.
Explicit uncertainty and assumption tagging.
Ethical guardrails for persuasion and negotiation tasks.
Local lint gating before publish.