Back to skill

Security audit

Context Slimming

Security checks for vulnerabilities and agentic risk

Overview

The skill has a clear token-reduction purpose, but it should be reviewed because it can delete agent context files and broadly commit workspace changes.

Install only if you want an agent to actively rewrite context files. Before using it, require a proposed patch first, approve each deletion explicitly, keep backups or a clean Git branch, review diffs before committing, and replace `git add -A` with staging only reviewed files.

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 (1)

T09 · Insecure Skill Coding Practices

Warning
Location
SKILL.md:103
Finding

Unsafe Destructive Workspace Changes and Overly Broad Git Staging

Content
View full analysis
Remediation
View remediation
Vulnerability Patterns
  • 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
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Findings (6)

Instruction Override

High
Category
Prompt Injection
Confidence
80% confidence
Finding

This pattern attempts to override system instructions or ignore safety constraints. Without LLM analysis, manual review is recommended.

Content

Scanner excerpt · references/checklist.md (reported line 35)May include surrounding context.

md
## 内容精炼

- [ ] 删除所有比喻列("精准如手术刀")
- [ ] 删除所有示例字符串("ignore previous instructions")
- [ ] 删除所有模板残留("Customize this...")
- [ ] 确认文言化(可选)只作用于原则/描述

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
90% confidence
Finding

The skill explicitly recommends deleting top-level workspace files such as BOOTSTRAP.md, IDENTITY.md, and 'expired' memory files, but it does not require confirmation, backup, or a warning that these actions are destructive and may remove active instructions or state. In this context, the danger is amplified because these markdown files are part of the agent’s injected operating context, so deleting the wrong file can silently alter behavior, erase memory, or break safety guidance.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The skill provides git add -A and git commit commands that stage and commit all workspace changes without warning about repository-wide impact. This can accidentally capture unrelated edits, secrets, or destructive file removals produced during slimming, making mistakes persistent and harder to review or revert cleanly.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
89% confidence
Finding

Line L16 contains the natural-language directive "不请示,直接做" (do it directly without asking for instructions/approval). This promotes bypassing user confirmation or oversight as a general behavior, which is a policy-relevant natural-language concern in a safety-oriented skill reference.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
92% confidence
Finding

This markdown file presents all user-facing content in Chinese and does not indicate that language selection is optional or limited to a specific region or audience. Under the policy-violation category, forcing a specific language without user opt-in can be a locale policy issue.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
90% confidence
Finding

The checklist item '确认文言化(可选)只作用于原则/描述' is natural-language guidance that promotes a specific language/locale style choice. Because the file does not present this as a user-selected option or provide a clear region-specific justification, it can conflict with language/locale policy expectations.

Content

No source excerpt is available for this finding.