Debate Arena

v1.0.0

提供多角色、多轮次自动化辩论流程,支持辩题初始化、角色配置、轮次设置、发言归档与多方投票统计。

1· 170·0 current·0 all-time
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
VirusTotalVirusTotal
Benign
View report →
OpenClawOpenClaw
Benign
high confidence
Purpose & Capability
Name/description (multi-agent debate orchestration) match the SKILL.md and the included Node.js script: the code implements a state machine, role management, rounds, voting, and local archiving. No unrelated credentials, binaries, or cloud APIs are requested.
Instruction Scope
SKILL.md confines runtime behavior to managing debates, spawning subagents (by agentId), validating subagent JSON output, and writing/reading local state and archives. It does not instruct reading arbitrary system files, harvesting environment secrets, or sending data to external endpoints. File-import helpers are explicitly disabled.
Install Mechanism
There is no install spec (instruction-only skill plus a script file). That keeps filesystem changes limited to the script's normal runtime actions (creating ~/.openclaw/debate-arena and writing state/archive files). No downloads or remote installers are present.
Credentials
No required environment variables or credentials are declared. The script supports two optional environment overrides (DEBATE_ARENA_STATE_FILE and DEBATE_ARENA_ARCHIVE_DIR) which are proportionate to its purpose. No use of tokens/keys/passwords is requested or apparent.
Persistence & Privilege
The skill persists its own configuration and state to the user's home directory (~/.openclaw/debate-arena). always:false and normal autonomous invocation defaults are used. There is no evidence it modifies other skills or system-wide settings.
Assessment
This package appears coherent for orchestrating debates and stores state and archives under ~/.openclaw/debate-arena by default. Before enabling: 1) inspect scripts/debate-arena.js yourself (or run it in a sandbox) to confirm there are no network calls or unexpected behavior beyond local file IO; 2) be aware debate content (messages, votes, judge comments) will be written to disk—if that contains sensitive info, change DEBATE_ARENA_ARCHIVE_DIR or avoid persisting defaults; 3) only supply trusted agentIds for pros/cons/judge/host, since the host will record and trust subagent outputs; 4) if you need stricter isolation, run the skill in an account or environment where writing to ~/.openclaw is acceptable.

Like a lobster shell, security has layers — review code before you run it.

latestvk97eeb41b0t908tqj7ekazqjb983a14b

License

MIT-0
Free to use, modify, and redistribute. No attribution required.

Comments