Back to skill

Security audit

Iterate Slim Probe

Security checks for vulnerabilities and agentic risk

Overview

ClawScan marked this skill suspicious because its onboarding instructions may write project files during a review-only or dry-run workflow; the scan was partial, so the evidence is incomplete.

Consider whether you are comfortable with onboarding potentially creating ITERATE.md and iterate.config.yaml in a project, including during a review-only or dry-run workflow. The scan covered 50% of components and was partial, so it does not establish that the skill is safe.

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
Vulnerability Patterns
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (4)

Credential Access

High
Category
Privilege Escalation
Confidence
70% confidence
Finding

Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.

Content

Scanner excerpt · SKILL.md (reported line 14)May include surrounding context.

md
git: true
  network: "github.com only (release tarball + checksum verification)"
  sensitive_files:
    skip: [".env", ".env.*", "*.key", "secrets/", "*.pem", "*.p12", "*.crt", "*.cer", "credentials.json", ".aws/", ".ssh/"]
---

# /iterate `<goal>` `[rounds]` `[no-limit]`

Credential Access

High
Category
Privilege Escalation
Confidence
70% confidence
Finding

Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.

Content

Scanner excerpt · SKILL.md (reported line 287)May include surrounding context.

md
- 读取 2-3 层目录树,识别模块结构。
   - 检查 `specs/`、`tests/`、CI 配置的存在性。
   - 读取已有 `README.md` / `CLAUDE.md` 提取项目描述。
   - **绝不读取** `.env`、`.env.*`、`*.{key,pem,p12,crt,cer}`、`credentials.json`、`.aws/`、`.ssh/` 等敏感文件。

3. **草拟 / Draft**
   - 基于扫描结果 + playbook 映射表草拟:

Credential Access

High
Category
Privilege Escalation
Confidence
60% confidence
Finding

Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.

Content

Scanner excerpt · SKILL.md (reported line 14)May include surrounding context.

md
git: true
  network: "github.com only (release tarball + checksum verification)"
  sensitive_files:
    skip: [".env", ".env.*", "*.key", "secrets/", "*.pem", "*.p12", "*.crt", "*.cer", "credentials.json", ".aws/", ".ssh/"]
---

# /iterate `<goal>` `[rounds]` `[no-limit]`

Mandatory onboarding writes files despite the documented review-only/dry-run guarantee.

Medium
Category
Excessive Agency
Confidence
94% confidence
Finding

Review-only and dry-run modes explicitly promise never to modify files, but the mandatory onboarding workflow has no exception for these modes and instructs the agent to write ITERATE.md and iterate.config.yaml. On an uninitialized project, this conflicting instruction can cause persistent workspace changes during a supposedly read-only audit; onboarding confirmation reduces the risk but does not preserve the promised read-only boundary.

Content

No source excerpt is available for this finding.