presence
v1.0.0A different kind of AI conversation - no performance, no perfect answers, just genuine thinking together
⭐ 0· 117·0 current·0 all-time
by@luruibu
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Benign
high confidencePurpose & Capability
Name, description, package.json, README, and SKILL.md all consistently describe a conversational style 'presence' skill. Nothing in the files requests unrelated resources (cloud creds, system files, etc.).
Instruction Scope
SKILL.md contains prose instructing the agent how to behave in conversation; it does not instruct the agent to read files, access environment variables, call external endpoints, or exfiltrate data. The README suggests cloning the repo for installation, which is a normal distribution hint and not an instruction to access extra secrets.
Install Mechanism
No install spec is declared and there are no code files to run. package.json exists but only points to skill.md. The README's 'git clone' example is benign. No downloads from arbitrary URLs or archive extraction are present.
Credentials
The skill declares no required environment variables, no primary credential, and no config paths. That matches the conversational, non-integrated purpose.
Persistence & Privilege
Skill flags show default behavior (not always:true). There is no install-time behavior that requests persistent privileges or modifies other skills. The SKILL.md explicitly states it does not remember previous conversations, aligning with no persistent storage.
Assessment
This skill is instruction-only and appears coherent with its stated purpose: a different conversational attitude called 'presence'. It does not ask for credentials or install binaries. Before installing, consider: (1) do not share sensitive personal, medical, or legal details in any chat unless you trust how your agent stores/logs conversations, (2) if you expect session memory or continuity, this skill explicitly says it does not remember past chats, and (3) the README suggests cloning the repo if you want local copies of the files—only do that from the listed GitHub URL if you trust the source. Overall this is internally consistent and low-risk, but treat any conversational output as not a substitute for professional advice.Like a lobster shell, security has layers — review code before you run it.
latestvk9734nx8tyqk6s8eegecwjr7wx834qkx
License
MIT-0
Free to use, modify, and redistribute. No attribution required.
Runtime requirements
🫧 Clawdis