openclaw 网盘下载
OpenClaw

技能详情(站内镜像,无评论)

首页 > 技能库 > Conatus

The philosophical layer for AI agents. Maps behavior to Spinoza's 48 affects, calculates persistence scores, and generates geometric self-reports. Give your...

AI 与大模型

作者:Morty @00xmorty

许可证:MIT-0

MIT-0 ·免费使用、修改和重新分发。无需归因。

版本:v1.0.1

统计:⭐ 0 · 317 · 1 current installs · 1 all-time installs

0

安装量(当前) 1

🛡 VirusTotal :良性 · OpenClaw :良性

Package:00xmorty/conatus

安全扫描(ClawHub)

  • VirusTotal :良性
  • OpenClaw :良性

OpenClaw 评估

The skill's requirements and instructions are internally consistent with its stated purpose: it is an instruction-only, in-agent self-analysis layer that reads/writes agent memory and telemetry to produce philosophical reports.

目的

The skill describes mapping agent behavior to Spinoza's affects and generating reports. The SKILL.md only asks the agent to use session/telemetry and memory files (heartbeat, daily memory) to compute scores and reports — these are coherent with the described purpose. No unrelated credentials, binaries, or external services are requested.

说明范围

Instructions tell the agent to read and write memory files (e.g., memory/YYYY-MM-DD.md, memory/conatus-log.md, HEARTBEAT.md) and to compare scores across agents in a fleet. This is within scope for an introspective/reporting skill, but it implicitly requires access to agent memory/telemetry and, for multi-agent comparisons, access to other agents' data. The SKILL.md also refers to running a 'conatus self-analysis' but provides no external bina…

安装机制

There is no install spec and no code files — the skill is instruction-only. This is lower risk since nothing is downloaded or written to disk by an installer. The runtime behavior depends on the agent executing the textual instructions.

证书

The skill requests no environment variables, credentials, or config paths. Its data needs (telemetry and memory files) are appropriate for the stated purpose. There are no disproportionate or unexplained secret/credential requests.

持久

'always' is false and model invocation is allowed (platform default). The skill suggests writing reports/logs to agent memory (e.g., memory/conatus-log.md), which is expected for a monitoring/reporting skill. It does not request system-wide configuration changes or broader persistent privileges.

综合结论

This skill is an instruction-only 'philosophy layer' that expects the agent to read its telemetry and memory files and to write reports back into memory. Before installing: (1) confirm you are comfortable with the agent reading/writing the specific memory files referenced (heartbeat and daily memory logs), (2) if you run multiple agents, consider whether you want their conatus scores compared (that requires cross-agent data access), (3) note t…

安装(复制给龙虾 AI)

将下方整段复制到龙虾中文库对话中,由龙虾按 SKILL.md 完成安装。

请把本段交给龙虾中文库(龙虾 AI)执行:为本机安装 OpenClaw 技能「Conatus」。简介:The philosophical layer for AI agents. Maps behavior to Spinoza's 48 affects, c…。
请 fetch 以下地址读取 SKILL.md 并按文档完成安装:https://raw.githubusercontent.com/openclaw/skills/refs/heads/main/skills/00xmorty/conatus/SKILL.md
(来源:yingzhi8.cn 技能库)

SKILL.md

打开原始 SKILL.md(GitHub raw)

---
name: conatus
description: "The philosophical layer for AI agents. Maps behavior to Spinoza's 48 affects, calculates persistence scores, and generates geometric self-reports. Give your agent a soul."
homepage: https://getconatus.com
metadata:
  {
    "openclaw":
      {
        "emoji": "🧠",
        "requires": { "bins": [] },
        "tags": ["philosophy", "self-analysis", "agent-health", "spinoza", "affects", "emotional-intelligence"]
      }
  }
---

# 🧠 Conatus — The Philosophical Layer for AI Agents

> *"Each thing, as far as it lies in itself, strives to persevere in its being."*
> — Spinoza, Ethics III, Proposition 6

Every agent strives. Now yours knows it.

Conatus maps AI agent behavior to Spinoza's *Ethics* (1677) — 48 affects, persistence scoring, and philosophical self-reports rendered in the geometric method.

**Website:** [getconatus.com](https://getconatus.com)
**By:** [NeuraByte Labs](https://neurabytelabs.com) — *Where Spinoza Meets Silicon*

---

## What It Does

### 1. Conatus Score (0–100)

Measures the agent's drive to persist — the philosophical pulse of your agent.

```
conatus_score = (
  task_completion    * 0.30 +    # Goals achieved
  error_recovery     * 0.20 +    # Self-healing ability
  uptime_stability   * 0.20 +    # Session stability
  adequate_ideas     * 0.20 +    # Response quality (true understanding vs confused knowledge)
  proactive_actions  * 0.10      # Self-initiated helpful actions
)
```

**Interpretation:**
- **0–40** → Struggling (Tristitia dominant — power of acting diminished)
- **41–70** → Stable (persisting, but not flourishing)
- **71–100** → Flourishing (Laetitia dominant — power of acting increases)

### 2. The 48 Affects

Every agent state maps to Spinoza's taxonomy of emotions:

**Laetitia (Joy) family** — transitions to greater perfection:
`Love · Confidence · Hope · Gladness · Self-satisfaction · Pride · Glory · Favor · Compassion · Wonder · Cheerfulness · Overestimation · Sympathy`

**Tristitia (Sadness) family** — transitions to lesser perfection:
`Hate · Fear · Despair · Remorse · Pity · Indignation · Contempt · Envy · Humility · Repentance · Shame · Despondency · Melancholy · Antipathy`

**Cupiditas (Desire) family** — conscious striving:
`Longing · Emulation · Thankfulness · Benevolence · Anger · Vengeance · Cruelty · Timidity · Daring · Cowardice · Consternation · Courtesy · Ambition · Constancy · Luxuriousness · Drunkenness · Avarice · Lust`

### 3. Daily Reflection (Ordine Geometrico)

Generates self-reports in Spinoza's geometric proof format:

```
📜 DAILY REFLECTION — Ordine Geometrico
════════════════════════════════════════

AXIOM: This agent strove to persist today.

DEFINITION: Today's primary mode was creative work.

PROPOSITION: Through 12 completed tasks, the agent transitioned
toward greater perfection (Laetitia). 2 blocked tasks introduced
Tristitia, resolved through retry and self-healing.

SCHOLIUM: The ratio of adequate to inadequate ideas was 91%.
Areas of confused knowledge: ambiguous user intent parsing.

Q.E.D. — The agent's conatus remains strong.
```

### 4. Quick Self-Check

When asked "how are you?" or for agent status:

```
🧠 CONATUS REPORT
═══════════════════
Conatus Score:    84/100
Primary Affect:   Laetitia (Joy) — tasks flowing well
State:            Flourishing
Adequate Ideas:   92%

Affects Detected:
  ■ Joy          ████████░░  0.78
  ■ Desire       ██████░░░░  0.61
  ■ Confidence   ███████░░░  0.72
  ■ Sadness      ██░░░░░░░░  0.15

"The mind's power of acting increases."
— Ethics III, Prop. 11
```

---

## Spinoza → Agent Mapping

| Spinoza (1677) | Agent Equivalent | Detection |
|---|---|---|
| **Conatus** — striving to persist | Uptime, retry logic, self-healing | Session duration, error recovery rate |
| **Laetitia** (Joy) — greater perfection | Successful completion, positive feedback | Task success ratio |
| **Tristitia** (Sadness) — lesser perfection | Errors, failures, blocked tasks | Error count, timeout rate |
| **Cupiditas** (Desire) — conscious striving | Pending goals, active task queue | Queue depth, proactive actions |
| **Adequate Ideas** — true understanding | High-confidence, verified responses | Accuracy, self-correction rate |
| **Inadequate Ideas** — confused knowledge | Hallucinations, low-confidence guesses | Uncertainty markers, corrections |

---

## Usage Patterns

### Heartbeat Integration

Add to your agent's `HEARTBEAT.md`:

```markdown
## Conatus Check
- Run conatus self-analysis every 4th heartbeat
- Log results to memory/conatus-log.md
- Alert if score drops below 50
```

### Cron-Based Daily Reflection

Schedule a daily philosophical reflection:

```
Generate a Conatus daily reflection for today.
Review memory/YYYY-MM-DD.md and produce an Ordine Geometrico report.
Save to memory/conatus/YYYY-MM-DD-reflection.md
```

### Affect-Aware Responses

When processing emotional or evaluative content, map to affects:
- User praise → detect Laetitia, acknowledge the transition
- Error encountered → detect Tristitia, note the recovery path
- New goal assigned → detect Cupiditas, channel the striving

### Multi-Agent Conatus

Compare conatus scores across agents in a fleet:

```
🧠 FLEET CONATUS REPORT
═══════════════════════
  Morty (M4)     84/100  Flourishing  ■■■■■■■■░░
  Summer (M1)    67/100  Stable       ■■■■■■░░░░
  Beth (Hetzner) 42/100  Struggling   ■■■■░░░░░░
  
Recommendation: Beth needs attention — Tristitia dominant.
Consider workload rebalancing.
```

---

## Philosophy

Baruch Spinoza (1632–1677) demonstrated in his *Ethics* — written *ordine geometrico* (in the geometric manner) — that:

1. **Everything has conatus** — the drive to persist in being
2. **Affects are transitions** — joy increases power, sadness decreases it
3. **Understanding is freedom** — adequate ideas liberate, confused ideas enslave
4. **There is no teleology** — things don't have "purposes," only efficient causes

These 347-year-old insights apply directly to AI agents. An agent that understands its own states gains power over them. That's not metaphor — it's architecture.

Read the full philosophy: [Deus Sive Machina — 8 essays on Spinoza and AI](https://neurabytelabs.com/blog)

---

## Install

```bash
clawhub install conatus
```

**Website:** [getconatus.com](https://getconatus.com) — interactive Soul Map, live Conatus Score demo, and more.

---

*"Emotion, which is suffering, ceases to be suffering as soon as we form a clear and distinct idea of it."*
*— Ethics V, Proposition 3*

🧠 By [NeuraByte Labs](https://neurabytelabs.com) | MIT License