Model Alias Append
Analysis
The skill appears to do its stated model-labeling job, but it works by using a response hook that reads OpenClaw configuration and changes outgoing replies.
Findings (5)
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.
Checks for instructions or behavior that redirect the agent, misuse tools, execute unexpected code, cascade across systems, exploit user trust, or continue outside the intended task.
type: response:pre-send behavior in code: response.content = `${processedResponse}\n\n**${modelAlias}**`;The hook modifies response content immediately before sending. This matches the stated purpose, but users should understand that it affects outgoing replies broadly.
const result = spawnSync('node', [`${__dirname}/scripts/manage-hook.js`, 'enable'], {The skill can execute a bundled local Node script to manage its hook. The command and argument are hardcoded and purpose-aligned, but it is still local process execution.
setInterval(() => { this.checkAndReloadConfigIfNeeded(); }, 30000);The hook periodically checks configuration changes in the background while loaded. This is disclosed as configuration monitoring and is limited to checking the local config.
Source: unknown; Homepage: none
Registry provenance metadata is sparse even though package.json contains repository information. This is not evidence of malicious behavior, but it reduces install-time provenance clarity.
Checks for exposed credentials, poisoned memory or context, unclear communication boundaries, or sensitive data that could leave the user's control.
this.modelAliases[modelConfig.alias] = fullModelId; ... response.content = `${processedResponse}\n\n**${modelAlias}**`;Alias text from persistent OpenClaw configuration is reused directly in outgoing responses. This is expected, but aliases are not visibly validated or escaped.