horus
v7.7.7Register, track, and validate state observations against declared constraints without assuming universal truth or enforcing physical laws.
⭐ 0· 672·0 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 files and SKILL.md present a protocol specification (CS-RV) and a metaphor-to-formal translation from a 'Horus' example; there are no unexpected environment variables, binaries, or install steps. The content matches a documentation/reference implementation skill.
Instruction Scope
SKILL.md contains explanatory text and reference Python/psuedocode for registering observations, declaring constraints, projecting states, and exporting traces. It does not instruct the agent to read arbitrary host files, require secrets, or transmit data to external endpoints. It does reference an optional external_endpoint URI (e.g., quantum://polymath.asi) and an EXPORT_TRACE operation that serializes state — these are placeholders in the spec and are not accompanied by concrete network-send instructions, but they are potential vectors if later code or a user runs the examples.
Install Mechanism
There is no install spec and no code files executed by an installer; this is the lowest-risk model (instruction-only).
Credentials
The skill declares no required environment variables, credentials, or config paths. The protocol mentions optional external endpoints in examples, but nothing requires secrets or access tokens.
Persistence & Privilege
The skill is not marked always:true and does not request persistent system privileges. It is user-invocable and allows autonomous invocation by default (the platform default), but it does not add additional persistence or attempt to change other skills' configurations.
Assessment
This appears to be a documentation/reference skill (protocol spec + example code) rather than executable tooling. It does not ask for secrets or install software, so it's low risk as-is. Before using it in an agent that can run code, review the example code and the companion skills.md to ensure there are no runtime instructions that would: (a) execute arbitrary code on your host, (b) POST exported traces to an external URI, or (c) require credentials. If you will let the agent invoke the skill autonomously and you are concerned about data leaving the agent, consider disabling autonomous invocation for this skill or verifying that EXPORT_TRACE outputs are stored locally and not transmitted to remote endpoints.Like a lobster shell, security has layers — review code before you run it.
arkhenvk974pabnghj8w9jyw9p5p1b1e180q9c5latestvk974pabnghj8w9jyw9p5p1b1e180q9c5
License
MIT-0
Free to use, modify, and redistribute. No attribution required.