Back to skill

Security audit

ia-writing-tests

Security checks for vulnerabilities and agentic risk

Overview

ClawScan rates this skill suspicious with medium confidence: it reports useful test-writing guidance, but flags instructions that could delete persistent data and a mismatch between the stated purpose and an evaluation tool described in the documentation.

Review the cleanup instructions carefully before installing or following them. Do not apply database or volume deletion steps unless you have verified the targets are disposable test resources and any important data is backed up. The scanner inspected all 12 listed components, but the implementation of the evaluation tool was not supplied, so its behavior remains uncertain.

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
  • Output HandlingUnvalidated Output Injection, Cross-Context Output, Unbounded Output
  • Memory PoisoningPersistent Context Injection, Context Window Stuffing, Memory Manipulation
  • 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
Findings (4)

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
88% confidence
Finding

The supplied commands target evaluation and diagnosis of ia-writing-tests, which is materially different from helping write or repair software tests. Nothing shown applies the described assertion-quality or anti-pattern guidance. The implementation of distiller.py is not supplied, so additional behavior or resource access cannot be established.

Content

No source excerpt is available for this finding.

Memory Manipulation

High
Category
Memory Poisoning
Confidence
80% confidence
Finding

Skill manipulates agent memory, state, or stored context. Memory corruption can alter personality, override safety rules, or cause unpredictable behavior.

Content

Scanner excerpt · references/test-smells.md (reported line 53)May include surrounding context.

md
**Symptom:** Methods like `reset()`, `clearState()`, `setTestMode()` that exist only because tests need them.

**Fix:** If tests need to reset state, the code has a design problem. Refactor to make state explicit and injectable.

### Snapshot tests as the only test

The fix ladder recommends destructive database and volume cleanup without an explicit data-loss warning.

Medium
Category
Output Handling
Confidence
94% confidence
Finding

The fix ladder recommends TRUNCATE, DROP DATABASE, and Docker volume deletion without requiring verification that the targets contain only disposable test data. Although infrastructure cleanup is relevant to testing, applying this guidance to a shared database or a Compose project with valuable persistent volumes could irreversibly delete unrelated data.

Content

No source excerpt is available for this finding.

The infrastructure-reset instructions recommend destructive cleanup without warning about data loss.

Medium
Category
Output Handling
Confidence
90% confidence
Finding

The reset guidance includes table truncation and volume teardown without explicitly limiting those operations to verified disposable test resources. The testing context makes the recommendation legitimate in purpose, but following it against shared or misconfigured infrastructure could destroy persistent data.

Content

No source excerpt is available for this finding.