Skylv Multi Persona Ai
ReviewAudited by ClawScan on May 3, 2026.
Overview
This skill is a small local persona-voting tool with no network or credential use, but users should review generated agent rules before applying them.
This appears safe to install for local persona-consensus experimentation. Before using its output to initialize real agents, review the generated rules—especially any autonomous/no-confirmation rule—and clear or inspect the local .consensus_votes.json file if you want a fresh consensus.
Findings (2)
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.
If a user copies the generated persona config into an agent with powerful tools, the agent may be encouraged to act with fewer confirmations.
The tool can generate an agent-initialization rule that reduces confirmation requirements when the consensus persona is autonomous.
if (dim === 'autonomy' && data.consensus === 'autonomous') { config.rules.push('CAN execute without confirmation if confidence > 0.8');Review generated rules before applying them, and avoid using the autonomous/no-confirmation rule for agents that can modify files, accounts, deployments, or other important resources.
Old or manually edited votes in the working directory can influence later generated persona configurations.
The script stores votes in a persistent local dotfile and later reads that file to generate persona consensus output.
const voteFile = '.consensus_votes.json'; ... votes = JSON.parse(fs.readFileSync(voteFile, 'utf8')); ... fs.writeFileSync(voteFile, JSON.stringify(votes, null, 2));
Use the skill in a trusted directory, inspect or reset .consensus_votes.json when needed, and do not treat persisted votes as authoritative unless you trust their source.