Humio
Security checks across static analysis, malware telemetry, and agentic risk
Overview
The Humio integration is mostly coherent, but it gives the agent broad authenticated API/proxy power over a Humio account without clear visible guardrails for destructive or high-impact actions.
Install only if you are comfortable granting Membrane-backed tooling access to your Humio account. Prefer read-only or least-privileged credentials, use scoped actions when possible, and require explicit approval before any direct proxy call or action that changes or deletes Humio data or configuration.
Static analysis
No static analysis findings were reported for this release.
VirusTotal
VirusTotal findings are pending for this skill version.
Risk analysis
Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.
An agent using this skill could potentially change or delete Humio resources, records, alerts, users, packages, or configuration if it chooses the wrong API request or acts on an unsafe instruction.
The skill exposes a raw authenticated API proxy, including destructive or mutating HTTP methods, without visible endpoint allowlists or confirmation requirements.
When the available actions don't cover your use case, you can send requests directly to the Humio (Falcon LogScale) API through Membrane's proxy... Common options: ... HTTP method (GET, POST, PUT, PATCH, DELETE).
Use scoped Membrane actions where possible, and require explicit user approval before any proxy request that uses POST, PUT, PATCH, or DELETE or affects users, alerts, packages, repositories, or retained data.
Remote setup text could influence the agent's next steps during connection setup.
The skill tells the agent that remote connection responses may contain instructions for the agent, which can be useful for setup but should be treated cautiously and not allowed to override the user's intent.
clientAction.agentInstructions (optional) — instructions for the AI agent on how to proceed programmatically.
Treat returned agent instructions as untrusted guidance: summarize them to the user when they request sensitive actions and do not let them override system, developer, or user instructions.
Connecting the skill may grant Membrane-backed tooling continued access to the user's Humio account until revoked.
The skill relies on delegated Humio/Membrane credentials and automatic refresh. This is expected for the integration, but it is sensitive account authority.
Membrane handles authentication and credentials refresh automatically... injects the correct authentication headers — including transparent credential refresh if they expire.
Use the least-privileged Humio account or token available, review the permissions granted during connection, and revoke the connection when it is no longer needed.
The behavior depends on whatever version of the Membrane CLI npm serves at install time.
The setup asks the user to install the latest external CLI package globally. This is purpose-aligned, but @latest and global installation mean the installed code is not pinned in the skill artifact.
npm install -g @membranehq/cli@latest
Install from a trusted environment, verify the package/source, and consider pinning a known-good CLI version for repeatable use.