ClawHub
Back to skill

Security audit

nudocs

Security checks for vulnerabilities and agentic risk

Overview

This skill coherently helps use the Nudocs CLI while explicitly warning about credentials, uploads, private links, overwrites, and deletion approvals.

Before installing, confirm you trust the Nudocs CLI package and understand that uploads send selected documents to Nudocs for processing. Use a protected local or CI secret for the API key, treat returned edit links as private, and require explicit confirmation for sensitive uploads, overwrites, public sharing, or deletion.

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
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • 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)

Autonomous Decision Making

Medium
Category
Excessive Agency
Confidence
80% confidence
Finding

Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.

Content

Scanner excerpt · SKILL.md (reported line 41)May include surrounding context.

md
- **Interactive personal setup:** if the current CLI exposes an interactive login/config flow, complete it in the user's terminal. Otherwise have the user enter the key directly in a protected local editor or secret manager, outside the agent transcript. Keep `~/.config/nudocs` mode `0700` and its API-key file mode `0600`.
- **Protected environment setup:** use `NUDOCS_API_KEY` only when it is already supplied through protected local or CI secret storage.

Never ask the user to paste the key into chat, put it in a command argument, echo it into a file, display it with `nudocs config`, or print environment/config contents. Verify authentication through behavior such as a bounded `nudocs list`, not by inspecting the secret.

## Discover the installed CLI contract

Autonomous Decision Making

Medium
Category
Excessive Agency
Confidence
75% confidence
Finding

Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.

Content

Scanner excerpt · SKILL.md (reported line 77)May include surrounding context.

nudocs pull 01ABC123XYZ --format md --output report.md

text

Resolve the exact output path first and do not overwrite an existing file without confirmation. Format conversion can lose layout, comments, or other features; do not promise lossless round trips unless the current service documents them.

Autonomous Decision Making

Medium
Category
Excessive Agency
Confidence
75% confidence
Finding

Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.

Content

Scanner excerpt · references/formats.md (reported line 41)May include surrounding context.

nudocs pull 01ABC123XYZ --format md --output report.md

text

Resolve the exact output path first and do not overwrite an existing file without confirmation. Format conversion can lose layout, comments, or other features; do not promise lossless round trips unless the current service documents them.

Static analysis

No suspicious patterns detected.