Learning Loop
v1.0.0Proactive learning engine that makes your agent smarter every day. Combines daily journaling, decision tracking with categories, lesson extraction with lifec...
⭐ 0· 145·0 current·0 all-time
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Benign
high confidencePurpose & Capability
The name/description (proactive learning, journaling, lesson extraction) matches the runtime instructions: creating a workspace tree, writing/reading journal, decision, lesson files, and running periodic reflections. It does not ask for unrelated credentials, external services, or binaries.
Instruction Scope
SKILL.md instructs the agent to create and operate on a local workspace (brain/, memory/) and to read/write files within that workspace (journals, decisions, lessons, inbox, archive). It does not instruct reading system config, home directories outside the workspace, environment variables, or exfiltrating data to external endpoints. The only minor ambiguity is how nightly runs are triggered (no scheduler described), but that is an operational gap, not a scope creep or secret access.
Install Mechanism
There is no install spec and no code files; the skill is instruction-only with zero dependencies. This is the lowest-risk install model and matches the claim 'Zero dependencies. Works immediately after install.'
Credentials
The skill declares no required environment variables, no credentials, and no config paths. The instructions only reference local workspace files that the skill itself creates — there are no disproportionate or unrelated secrets requested.
Persistence & Privilege
always:false (default) and autonomous invocation is allowed (also default). The skill does not request forced permanent presence or modify other skills. It writes its own workspace files, which is expected for this purpose.
Assessment
This skill appears to do what it says: manage local journaling, decision-tracking, and lesson files. Before installing, decide and control where the workspace lives — do not point it at /, your whole home directory, or other sensitive folders. Expect the agent to write potentially sensitive content (journals may contain PII, credentials, or proprietary info); keep that workspace access-restricted and back it up appropriately. Verify how nightly reflections are triggered in your agent runtime (manual invocation vs scheduler) so it doesn't run unexpectedly. Confirm the agent enforces the 'Never overwrite existing files' rule and that promotions to permanent rules require explicit user confirmation (the SKILL.md describes asking the user, but verify your agent prompts and records that interaction). Finally, if you want remote backups or integrations, review and restrict any added connectors before enabling them.Like a lobster shell, security has layers — review code before you run it.
latestvk976zg9mpzsfr1dcjm6rzerx8d83tyzc
License
MIT-0
Free to use, modify, and redistribute. No attribution required.
Runtime requirements
🔄 Clawdis
OSLinux · macOS · Windows