Learning
v1.0.0Auto-learns how you learn best. Adapts teaching style, format, and depth to you.
⭐ 7· 5.4k·62 current·68 all-time
byIván@ivangdavila
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Benign
high confidencePurpose & Capability
Name/description match the implementation: SKILL.md + reference docs define how the agent should detect and record learning preferences. No binaries, env vars, or installs are requested, which is appropriate for an in-agent personalization skill.
Instruction Scope
Instructions tell the agent to 'detect patterns' and update SKILL.md using guidance in criteria.md and dimensions.md. That is within scope, but somewhat vague about the exact signals and data sources (conversation content, metadata, etc.). The skill will persist preferences by editing its own SKILL.md — expected for personalization but worth knowing.
Install Mechanism
No install spec and no code files — lowest-risk model. Nothing is downloaded or written beyond the SKILL.md updates the skill itself is designed to make.
Credentials
The skill requests no environment variables, credentials, or config paths, which is proportional to its described purpose.
Persistence & Privilege
The skill implicitly persists preferences by editing SKILL.md (allowed behavior). It is not always-on and does not request elevated system-wide config changes. If you care about privacy, be aware preferences will be stored in the skill's files unless the agent is instructed otherwise.
Assessment
This skill appears to do what it says and doesn't request external credentials or installs. Before enabling: (1) understand that the agent will observe conversation content and write compact preferences into the skill's files (SKILL.md) — review where those files are stored and who can access them; (2) if you prefer to approve changes, require the agent to ask before updating preferences or disable autonomous invocation for this skill; (3) periodically inspect SKILL.md for any sensitive text that might have been saved accidentally; and (4) if you need stricter privacy, avoid using the skill or keep it user-invocable only so it acts only when you ask.Like a lobster shell, security has layers — review code before you run it.
latestvk974acr3416zgx4qy9qb0s03qs80ysyf
License
MIT-0
Free to use, modify, and redistribute. No attribution required.