Real Skill packageSource verifiedClawHub registry

adversarial-debate

Cross-vendor adversarial review. Ship a plan, proposal, or design to a model from a DIFFERENT vendor to attack it; every objection carries a verifiable anchor; the defender rules with an evidence tag on each ruling; the final round classifies into still-disputed / unresolved / verified-consensus instead of forcing agreement; a fresh-session judge is mandatory whenever the outcome looks too clean. Invoke only when the user explicitly asks for an adversarial review by a model from another vendor. One model role-play…

Identity and source

Publisher attributionXiaobaregistry owner unverified by skillvetai
Functional categoryImage & Visual Designautomatically inferred · 56% rule confidence
Package forminstruction bundle2 recorded files
Canonical sourceClawHub registryclawhub:xiaoba-dev:adversarial-debate
Open canonical source ↗

Platform declarations

These states come from the source or distribution context. None of the entries below are SkillVetAI compatibility test results.

OpenClawnative officialProvenance: registry distribution

Independent structural checks

These checks parse the fixed package against dated platform rules. They do not execute the Skill or verify task behavior.

Claude Codepasses structure
Checker 0.1.0 · agent-skills-2026-08-13+claude-code-docs-2026-08-13 · 8/30/2026.claude/skills/adversarial-debate

Runtime, accounts, dependencies, permissions, network behavior and task quality remain untested.

OpenAI Codexpasses structure
Checker 0.1.0 · agent-skills-2026-08-13+codex-docs-2026-08-13 · 8/30/2026.agents/skills/adversarial-debate

Runtime, accounts, dependencies, permissions, network behavior and task quality remain untested.

OpenClawpasses structure
Checker 0.1.0 · agent-skills-2026-08-13+openclaw-docs-2026-08-13 · 8/30/2026skills/adversarial-debate

Runtime, accounts, dependencies, permissions, network behavior and task quality remain untested.

Installation and inspection

This command is recorded from the source ecosystem and resolves the registry's latest release. The fixed release shown on this page should be inspected before adoption.

clawhub install @xiaoba-dev/adversarial-debate
clawhub inspect @xiaoba-dev/adversarial-debate --version 2.2.0

Security evidence

SkillVetAI static result: no findings detected

This automated, non-executing scan is bound to this release hash. It is not a safety certification and may contain false positives or false negatives.

Status
completed
Coverage
full text content
Files
2 / 2 inspected as text
Checked
8/30/2026, 4:44:55 PM
Scanner
0.1.3
Policy
1.0.3
2 inferred permission indicators
  • network access — automatically inferred
  • filesystem write — automatically inferred
1 dependency and API indicators
  • api: clawhub.ai
External clawhub result: clean

This is registry-supplied evidence for the recorded release, not an independent SkillVetAI scan. Check the canonical source for the full report, scanner versions, scope, and current moderation state.

Evidence checked
8/30/2026, 2:37:50 PM
Release binding
Matches this record
  • vt: clean
  • skillspector: clean
  • llm: clean

Recorded files

The catalog stores hashes and an inventory summary for change detection. It does not republish the package contents.

Package content hashsha256:57460495451bcd3b746c7c3a0b12cb6e49a5b6cf8114be1eb69472c86af2e7f2
Show up to 2 recorded paths
  • skill-card.md
  • SKILL.md

Source changelog

Methodology upgrade, not just wording. R1 objections now require a verifiable anchor (file:line, runnable command, citable data) plus a mandatory closing question: what's the failure scenario for NOT acting. R2 rulings carry an evidence tag - verified / reasoned / to-verify - and only verified items are consensus-eligible. R3 no longer forces a consensus list: it outputs three categories (still disputed / unresolved-no-evidence-either-side / verified consensus), and failing to converge is not treated as a failure. R4 fresh judge is now mandatory whenever a trigger fires (a prohibition in the consensus, zero rejections all session, or 5+ consensus items) instead of optional. Core stances gained a hard number: across an archived sample, only ~1 in 40 objections was actually rejected, and every real rejection came from running production data - a warning against mistaking agreement for a w…

Release security diff

mediumCompared fixed releases 2.1.1 and 2.2.0; 0 finding and 0 permission indicators were added.

Both fixed releases were scanned under the current scanner and policy, so finding, permission and dependency changes are available.

Change reasons and limitations
  • content hash changed

Observed release history

These older immutable releases were observed by prior successful syncs. They remain recorded when a newer release becomes current.

2.1.18/29/2026sha256:a3d6bbd14f0db26ebbe1e7bbf895acbb4f703caa768e52619fb97106aab222cc
2.1.08/28/2026sha256:753f5ea5f7d7c0fe0284e97a818862085efe2c0ef85415889e40077f8596f940