Nm Conserve Context Optimization

• Outcome: Pass
• Latest audit: Jun 7, 2026 at 9:43 PM UTC
• Version:
  1.0.3

Overview

This is a documentation-only context-management skill with purpose-aligned guidance, but users should be careful with its checkpoint examples because they may write sensitive working state to disk.

Install only if you want an agent to use context-conservation and subagent-coordination patterns. Avoid using its checkpointing examples with secrets, credentials, private prompts, or sensitive intermediate results unless you change them to use a private per-run path, tight permissions, minimal structured state, and cleanup.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Rogue Agent: Self-Modification, Session Persistence
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access
• Supply Chain: Unpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (2)

Session Persistence

Medium

• Category: Rogue Agent

Content

# Pattern: Externalize critical state before compaction risk
def preserve_subagent_state(progress):
    """
    Write state to files so it survives compaction.
    """
    # Write to TodoWrite for task state
    todo_state = {

• Confidence: 80% confidence
• Finding: Write state to file

Session Persistence

Medium

• Category: Rogue Agent

Content

# Pattern: Externalize critical state before compaction risk
def preserve_subagent_state(progress):
    """
    Write state to files so it survives compaction.
    """
    # Write to TodoWrite for task state
    todo_state = {

• Confidence: 80% confidence
• Finding: Write state to files so it survives compaction. """ # Write to TodoWrite for task state todo_state = { 'completed': progress.completed_tasks, 'pending': progress.pending_ta

VirusTotal

65/65 vendors flagged this skill as clean.