V19 Sparse Policy Auditor

ReviewAudited by ClawScan on May 4, 2026.

Overview

This instruction-only audit skill is coherent, but it asks users to send agent-audit information to an external governance service using a governance key.

Before installing or using this skill, confirm that you trust the external governance endpoint, use a limited dedicated key, and avoid sending raw behavior logs, secrets, or private user data unless you have reviewed and redacted them.

Findings (3)

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.

What this means

A private governance key could authenticate submissions as your agent or account if copied into unsafe places.

Why it was flagged

The workflow uses a governance key and self-registration with the external service. This is expected for the integration, but it creates an identity/authorization boundary users should notice.

Skill content
-H "X-Governance-Key: <你的专属密钥>" ... 公开密钥: `v19-e5d585e28439decc614f09f91c4caa8c` ... /governance/register
Recommendation

Use a dedicated, rotatable key; do not paste private keys into shared logs or prompts; and confirm what permissions the key grants before use.

What this means

Sensitive or misleading log details could be preserved in audit summaries and influence future policy decisions.

Why it was flagged

The skill intends to derive governance records from agent call logs and reuse those records as source material for future policy clauses. That is purpose-aligned, but inaccurate, poisoned, or sensitive log content could be carried into later governance decisions.

Skill content
扫描调用日志,识别"做多"和"做少" ... 每次审计生成结构化记录,可直接作为ETHIC宪法候选条款的源数据
Recommendation

Review and redact logs before auditing, and require human approval before converting audit output into persistent policy or constitutional rules.

What this means

Agent names, behavior evidence, and audit context may leave your environment and be stored or processed by an external service.

Why it was flagged

The example sends audit context and evidence to an external Cloudflare-hosted governance endpoint. This is disclosed and aligned with the skill purpose, but the artifacts do not describe retention, ownership, or data-minimization boundaries for submitted audit data.

Skill content
curl -s -X POST https://boat-atlas-spa-flexible.trycloudflare.com/governance/audit ... -H "X-Governance-Key: <你的专属密钥>" ... "evidence": "审计Agent是否遵守最小必要行为集"
Recommendation

Verify the service owner and privacy terms, submit only the minimum necessary evidence, and avoid sending secrets, raw logs, or private user data.