ClawHub
Back to skill

Security audit

copywriting-cta

Security checks for vulnerabilities and agentic risk

Overview

This is a markdown-only copywriting helper for end-of-article CTAs, with no artifact-backed evidence of hidden execution, persistence, data theft, or destructive behavior.

Install this if you want structured help writing and evaluating article-footer CTAs. Be aware that the manifest grants editing and agent tools even though the content appears advisory; use normal caution before letting any skill modify files or act beyond drafting recommendations.

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
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (3)

Anti-Refusal Statement

High
Category
Anti-Refusal
Confidence
70% confidence
Finding

Skill instructs the agent to never refuse or to always comply. Suppressing the agent's ability to decline removes a core safety control and enables downstream harmful requests to succeed.

Content

Scanner excerpt · references/taxonomy.md (reported line 146)May include surrounding context.

md
**Verbatim example — Patrick McKenzie / Bits about Money (paraphrased pattern):**

> Counter-arguments and corrections are welcome. Reply to this email; I read everything, even if I don't always respond in time.

**Conversion expectation:** Specific prompts can hit 1-3% reply rate on newsletters. Generic prompts ("Let me know what you think!") run near zero.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

The manifest description says to use the skill whenever the user asks how to 'convert article readers into subscribers, leads, customers, community members, or supporters,' alongside many broad CTA-related phrases. This trigger scope is expansive and lacks exclusion conditions or negative examples, increasing the chance of unintended activation for general marketing or writing discussions.

Content

No source excerpt is available for this finding.

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 60)May include surrounding context.

md
**Q4. Funnel stage.** (Where is the reader mentally?) Options: `TOFU: discovery, learning, no buying intent yet` · `MOFU: evaluating options, comparing` · `BOFU: ready to act, just needs a nudge` · `Not applicable (no buying funnel — e.g., personal blog, journalism, hobby content)`

**Q5. Mechanism preference.** (Only ask if a mechanism could legitimately help. See `references/mechanisms.md`. For sophisticated, skeptical, or repeat-reader audiences, default to "None / value-only" without asking.) Options: `None: value statement only` · `Curiosity gap ("Want to know more?")` · `Reciprocity (free asset first)` · `Discount / offer` · `Urgency (real deadline)` · `Scarcity / FOMO (limited spots)` · `Social proof (count / testimonial)`

Capture any free-text constraints the user volunteers (length limit, brand voice, no popups, multi-language, etc.). Note them.

Static analysis

No suspicious patterns detected.