Deterministic persona reputation engine that applies guard decision effects to persona_set state and emits explicit reputation_delta artifacts.

This skill is coherent and small: it only needs Node and processes JSON input to compute reputation deltas. You can safely inspect the included src/index.mjs (it contains the full logic) before installing. Note the default ruleset (reward/penalty values and min/max reputation) is used unless you supply a ruleset; review or override those if they must match your policy. Because it can be invoked by agents, decide whether you want to allow autonomous runs in your environment, but there are no network calls or secret accesses in the code.

Type: OpenClaw Skill Name: consensus-persona-engine Version: 0.1.1 The OpenClaw skill 'consensus-persona-engine' is a deterministic persona reputation engine. It processes structured JSON input to update persona reputations based on predefined rules and emits reputation delta artifacts. The code primarily performs computational logic, uses `node:fs` to read its designated input file, and `node:crypto` for UUID generation. There is no evidence of data exfiltration, malicious execution, persistence mechanisms, prompt injection attempts against the agent, or obfuscation. The use of `fs.readFileSync` is for its stated purpose of reading an input file and does not indicate malicious intent within the skill itself.