Core Refinery
v1.0.5Find the core that runs through everything — the ideas that survive across all your sources.
⭐ 2· 1.5k·3 current·3 all-time
byLee Brown@leegitw
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Benign
high confidencePurpose & Capability
Name/description (finding invariants across multiple sources) match the SKILL.md instructions. The skill expects 3+ sources and/or extractions from related skills (essence-distiller, pattern-finder) which is reasonable for this use case.
Instruction Scope
The SKILL.md confines processing to the agent's configured model and explicitly says it does not write files. It asks for raw texts or extractions and describes a clear N-count synthesis process. Note: it will process whatever sources you provide (including confidential content) via whatever model the agent is configured to use; if that model is cloud-hosted, your data will be sent to the model provider as part of normal operation.
Install Mechanism
Instruction-only skill with no install spec and no code files. No artifacts are downloaded or written to disk by the skill itself.
Credentials
Requires no environment variables, credentials, or config paths. Declared inputs (3+ sources or extractions) are proportional to the stated task.
Persistence & Privilege
always:false and user-invocable:true. The skill does not request persistent presence or system-wide configuration changes.
Assessment
This skill appears internally consistent and lightweight. Before using it, be mindful that: (1) it will process whatever source texts you give it using the agent's configured model — if that model is cloud-hosted, data will be sent to the provider, so avoid feeding highly sensitive or proprietary content unless your privacy policy allows it; (2) the skill expects structured extractions or raw sources and may rely on outputs from other skills — ensure those upstream skills don't broaden access to secrets; (3) Golden Master candidates are heuristic (N≥3) — verify and vet candidates before publishing or operationalizing them. If you need stricter data residency, run the agent with an on-premise model or scrub/redact sensitive fields before submitting sources.Like a lobster shell, security has layers — review code before you run it.
analysisvk97d02p3cb5earcx4f186y6se183jhhqconsolidationvk97d02p3cb5earcx4f186y6se183jhhqknowledge-managementvk97d02p3cb5earcx4f186y6se183jhhqlatestvk97d02p3cb5earcx4f186y6se183jhhqmergingvk97d02p3cb5earcx4f186y6se183jhhqmulti-sourcevk97d02p3cb5earcx4f186y6se183jhhqopenclawvk97d02p3cb5earcx4f186y6se183jhhqsummarizationvk97d02p3cb5earcx4f186y6se183jhhqsynthesisvk97d02p3cb5earcx4f186y6se183jhhq
License
MIT-0
Free to use, modify, and redistribute. No attribution required.