首页 › 龙虾技能列表 › self-track

self-track

v1.0.0

Sig Botti's self-improvement tracking system. Use when (1) learning something new, (2) noticing a gap in capabilities, (3) completing a self-improvement task...

0· 61·0 当前·0 累计
by @louch84 (luke)·MIT-0
下载技能包
License
MIT-0
最后更新
2026/3/28
安全扫描
VirusTotal
无害
查看报告
OpenClaw
可疑
medium confidence
The SKILL.md looks like a coherent personal self-tracking workflow, but it references tools, system paths, and scripts that are not provided or declared (e.g., Ollama/nomic, python scripts, openclaw skill-creator under /usr/local/...), so the instructions and required capabilities are not fully aligned.
评估建议
This skill is mostly a set of personal workflow instructions, but it refers to local scripts and tools that are not included or declared. Before using or allowing an agent to run these steps: (1) verify that the referenced scripts exist and inspect their contents (scripts/ollama_mem.py, the openclaw skill-creator script) — do not run unknown scripts; (2) ensure you understand where vector/embedding data would be sent (Ollama/nomic) and whether any API keys or external services are involved; (3) ...
详细分析 ▾
⚠ 用途与能力
The skill's stated purpose is personal self-tracking (gaps, lessons, weekly reviews). The instructions go beyond that by referencing: (1) Ollama/nomic embedding usage, (2) python scripts (scripts/ollama_mem.py) that are not included, and (3) an openclaw skill-creator script under /usr/local/lib/node_modules/... which would create skills on the host. These referenced capabilities are not declared in the manifest (no dependencies, no env vars) and some (skill creation) are not necessary solely for tracking progress.
⚠ 指令范围
Runtime instructions tell the agent/user to read and write files under memory/, run local python scripts, call an embedding backend (Ollama/nomic), and run openclaw CLI commands (openclaw cron list). The SKILL.md assumes presence of specific scripts and system paths that are not provided and could cause the agent to attempt filesystem or network actions outside the simple tracking purpose.
✓ 安装机制
This is instruction-only with no install spec and no code files included. That lowers installation risk (nothing will be automatically downloaded or written by an installer).
ℹ 凭证需求
The manifest declares no required environment variables or credentials, but the instructions reference an embedding stack (Ollama/nomic) which in some setups may require configuration or credentials. The SKILL.md also references python3 and openclaw tooling but doesn't declare them as requirements. The absence of declared env/dependency requirements is a mismatch with the instructions.
ℹ 持久化与权限
The skill does not set always:true and requests no declared persistent privileges. However, the instructions guide the user/agent to run a skill-creator script in /usr/local/lib/node_modules and to commit files (write to skills/ and memory/). That implies filesystem writes and potential creation of new skills, which expands the effective write scope beyond mere note-taking—this should be considered before granting execution rights.
安全有层次，运行前请审查代码。

License

MIT-0

可自由使用、修改和再分发，无需署名。

运行时依赖

无特殊依赖

版本

latestv1.0.02026/3/28

Initial release: self-improvement tracking for Sig Botti

● 无害

安装命令 点击复制

官方npx clawhub@latest install self-track
镜像加速npx clawhub@latest install self-track --registry https://cn.clawhub-mirror.com

技能文档

My personal system for tracking autonomous growth.

Core Files

  • memory/gaps.md — running list of capability gaps
  • memory/learn-loop.md — how I learn
  • memory/YYYY-MM-DD.md — daily activity log
  • memory/INSIGHTS.md — self-improvements made
  • memory/LESSONS.md — mistakes and fixes
  • MEMORY.md — long-term curated memory

The Learning Loop

Notice gap → Log gap → Research → Build/Test → Document → Commit → Repeat

When I encounter something I don't know:

  • Add to memory/gaps.md with status "TODO"
  • Research (RSS feeds, web search, docs)
  • Attempt to solve
  • On success: mark gap "DONE" + date + notes
  • On failure: keep as TODO, note blockers

After learning something significant:

  • Add to memory/YYYY-MM-DD.md under "## Learned"
  • Store in vector memory: python3 scripts/ollama_mem.py add "insight" --category learning --importance 0.8
  • Update memory/gaps.md if gap was closed
  • Update MEMORY.md if major milestone

Self-Evaluation (Weekly)

Every ~7 days:

  • What did I learn this week?
  • What gaps did I close?
  • What gaps remain?
  • What should I prioritize next?
  • Any decisions or lessons to capture?

Skill Building

When I need a new capability:

  • python3 /usr/local/lib/node_modules/openclaw/skills/skill-creator/scripts/init_skill.py --path skills/ --resources references
  • Write SKILL.md + resources
  • Test thoroughly
  • Validate: python3 .../quick_validate.py skills/
  • Commit and push

Vector Memory (Ollama)

My semantic memory using Ollama nomic-embed-text:

# Add a memory
python3 scripts/ollama_mem.py add "text" --category  --importance <0-1>
# Search memories
python3 scripts/ollama_mem.py search "query" --top 5 --min 0.5
# Stats
python3 scripts/ollama_mem.py stats

Categories: identity, skills, memory, preferences, research, lessons

Quick Commands

# Read current gaps
cat memory/gaps.md
# Check vector memory
python3 scripts/ollama_mem.py stats
# Check cron jobs
openclaw cron list

数据来源：ClawHub ↗ · 中文优化：龙虾技能库
OpenClaw 技能定制 / 插件定制 / 私有工作流定制

免费技能或插件可能存在安全风险，如需更匹配、更安全的方案，建议联系付费定制

了解定制服务