Back to skill

Security audit

TaskFlow-ClawHub Publish 编排技能

Security checks for vulnerabilities and agentic risk

Overview

The skill’s publishing purpose is clear, but it can perform an irreversible external publish without a clear human approval gate and includes unsafe reference patterns for constructing child-agent commands.

Review this skill carefully before installing or using it for real publishing. Only run it with trusted, reviewed skill folders and sanitized slug/name/version/changelog values, and require an explicit manual approval immediately before the live publish command. Prefer structured argument passing over child-agent command strings, and avoid the node -e slug interpolation pattern in the verification script.

Vulnerability Patterns
  • Insecure Skill Coding PracticesFinds exploitable flaws such as hardcoded secrets or command injection
  • 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
Findings (2)

T09 · Insecure Skill Coding Practices

Error
Location
references/flow.ts:13
Finding

Command and Instruction Injection Through Unquoted Child-Task Construction

Content
View full analysis
Remediation
View remediation

T09 · Insecure Skill Coding Practices

Error
Location
references/inspect.sh:19
Finding

JavaScript Code Injection Through Slug Interpolation in node -e

Content
View full analysis
Remediation
View remediation
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (5)

Tp4

High
Category
MCP Tool Poisoning
Confidence
95% confidence
Finding

The code generally matches the stated orchestration and publishing purpose: it uses TaskFlow to create a managed flow, performs a dry-run validation command, executes a publish command to ClawHub, and records final state. However, the description explicitly says the flow validates, packages, publishes, and then verifies the published artifact. In the provided code, there is no separate packaging action and no post-publication verification step. The flow is finished immediately after the publish task is launched/recorded, with no evidence of artifact verification. Therefore the description overstates the implemented behavior in material ways.

Content

No source excerpt is available for this finding.

Ae1

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 30)May include surrounding context.

md
contains a valid `SKILL.md` frontmatter block.

Ae1

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 42)May include surrounding context.

md
contains a valid `SKILL.md` frontmatter block.

Ae1

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 65)May include surrounding context.

md
for the verification commands, and `references/flow.ts` for the TaskFlow

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
89% confidence
Finding

The skill describes performing a live, irreversible publish to an external registry but does not prominently require explicit user confirmation immediately before the write. In agentic contexts, this increases the risk of unintended publication, supply-chain mistakes, or accidental release of sensitive or unreviewed content because the action cannot be cleanly rolled back.

Content

No source excerpt is available for this finding.