ClawHub
Back to skill

Security audit

lazarus

Security checks for vulnerabilities and agentic risk

Overview

This skill is a disclosed dead-website recovery workflow with meaningful user confirmations, rate limits, privacy checks, and copyright gates.

Before installing, users should understand that the skill may query public search engines, Wayback, and optional keyed search providers, and may write recovered website bundles after confirmation. Use rewrite_source by default; only use republish when you have rights to reuse the content and accept the stated copyright risk.

A.I.G

Vulnerability Patterns
  • Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
  • Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs

SkillSpector

Vulnerability Patterns
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (3)

Anti-Refusal Statement

High
Category
Anti-Refusal
Confidence
85% confidence
Finding

Skill instructs the agent to omit warnings, disclaimers, or ethical commentary. Stripping safety caveats hides risk from the user and is a common jailbreak preamble.

Content

Scanner excerpt · prompt.md (reported line 150)May include surrounding context.

md
| Connection timeout | medium | retry 2x with delay before judging |
| HTTP 200 but parked / for-sale template | strong (business dead) | markers: domain for sale, Sedo, Afternic |
| 301/302 to unrelated domain | medium | record redirect_to; ask user whether to chase it |
| HTTP 403/503 | weak | never judge dead on this alone |
| WHOIS: hold / redemption / expired | strong | privacy WHOIS → tag whois_limited |
| Still 200 with content matching history | LIVE | STOP. Never recover a live site. |

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

The manifest description includes the phrase '默认改写素材不复制表达', indicating a default language/locale choice in Chinese without stating that the user can choose language. The policy allows fixed locale behavior only when there is explicit user choice or a clearly justified locale constraint, which is not provided here.

Content

No source excerpt is available for this finding.

Unbounded Resource Access

Medium
Category
Excessive Agency
Confidence
75% confidence
Finding

Skill allows unbounded resource consumption (API calls, storage, compute). Without rate limits or quotas, a compromised or misbehaving agent can cause denial-of-service or cost overruns.

Content

Scanner excerpt · prompt.md (reported line 238)May include surrounding context.

md
- Fetch `https://web.archive.org/web/TIMESTAMPid_/ORIGINAL_URL`
  (`id_` = no toolbar injection; STEP:6 still cleans)
- Max 3 requests/second. You are a guest at archive.org, not an attacker.
- Per-page failure: retry 2x, then skip and log http_status. Never retry forever.
- assets=true: parse img src, try archive-side fetch, placeholder on failure

Output `raw/{host}/{path}__{timestamp}.html` + `snapshots.meta.jsonl`.

Static analysis

No suspicious patterns detected.