OpenClaw Knowledge Runtime
v1.0.2Build a standalone layered knowledge runtime with typed links across knowledge entries, entities, memories, and reusable assets. Use when designing or implem...
⭐ 0· 187·4 current·4 all-time
byWANGJUNJIE@wanng-ide
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 actual contents: documentation and a tiny index.js implement a standalone knowledge/runtime design. Required artifacts (storage files, adapters, record types) align with the stated goal; nothing asks for unrelated cloud credentials or external services.
Instruction Scope
SKILL.md confines activity to building queries, reading layered stores, one‑hop link expansion, ranking, and writing back only stable findings. It references default local storage paths under memory/knowledge/*. This is appropriate for a local runtime, but the instructions imply the agent will read and (occasionally) write files in those locations when integrated — verify host adapters constrain what gets read/written and avoid dumping raw logs or sensitive data into the store.
Install Mechanism
No install spec. The skill is instruction-only plus a small index.js and docs; nothing downloads or executes external installers or archives.
Credentials
No required environment variables, credentials, or config paths are declared. The runtime uses only local file paths and adapter interfaces as described — proportional to the stated functionality.
Persistence & Privilege
always is false and the skill does not request persistent platform privileges. It describes write-back behavior but leaves actual persistence to the host through adapters; no self-enabling or cross-skill configuration changes are present.
Assessment
This skill appears coherent and local-only, but before installing: (1) confirm where the host will place and permission the memory/knowledge/* files so sensitive information isn't written to an insecure location; (2) review any host adapter code that integrates this runtime (query_builder, retrieval_selector, write_back) to ensure it filters sensitive inputs and enforces write-back rules; (3) when deploying in production, enforce retention/rotation and access controls on the knowledge store and audit what the skill writes to avoid accidentally persisting raw logs or secrets.Like a lobster shell, security has layers — review code before you run it.
latestvk970n37bv8nn4q00pd31brzbwx834rc6
License
MIT-0
Free to use, modify, and redistribute. No attribution required.