Career Spotlight Finder

• Outcome: Pass
• Latest audit: Sep 15, 2026 · 1:36 AM UTC
• Version: 1.0.2

Overview

The skill is coherent for career-branding analysis, but it deliberately saves local career files and should be used with trusted inputs.

Before installing, know that this skill saves analyses, reports, copy, and old versions under ~/.career-spotlight/. Use ClawHub or a pinned installer where possible, provide only projects and URLs you trust, and delete or move the local profile directory if you do not want those career materials retained.

A.I.G

Vulnerability Patterns

• Insecure Dependencies: Introduces malicious components through unsafe dependency sources
• Insecure Skill Coding Practices: Finds exploitable flaws such as hardcoded secrets or command injection
• Skill Instruction Hijacking: Alters the agent's session goals or safety constraints when the skill loads
• Agent Memory Poisoning: Writes attacker-controlled rules into memory that affect later sessions
• Remote Payload Retrieval and Execution: Fetches external code whose behavior can change after review

Findings (2)

T08 · Insecure Dependencies

Warning

• Location: README.md:18

Finding

Unpinned Package Execution During Installation

Content

View full analysis

Remediation

View remediation

add RealZST/career-spotlight-finder ``` 2. Document the expected package publisher, version, and integrity digest. 3. Prefer installation from an immutable Git commit or signed release rather than a mutable branch or latest package release. 4. Recommend reviewing downloaded content before enabling the Skill. 5. Use lockfiles and integrity verification wherever the installation mechanism supports them. 6. Publish signed release tags and document how users can verify signatures or checksums. ]]>

T09 · Insecure Skill Coding Practices

Warning

• Location: guides/input-collection-guide.md:42

Finding

Insufficient URL Validation Permits Server-Side Request Forgery

Content

View full analysis

Remediation

View remediation

SkillSpector

Vulnerability Patterns

• Supply Chain: Unpinned Dependencies, External Script Fetching, Obfuscated Code
• Excessive Agency: Unrestricted Tool Access, Autonomous Decision Making, Scope Creep
• Rogue Agent: Self-Modification, Session Persistence
• Trigger Abuse: Overly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
• MCP Tool Poisoning: Hidden Instructions, Unicode Deception, Parameter Description Injection

Findings (26)

Session Persistence

Medium

• Category: Rogue Agent
• Confidence: 60% confidence

Finding

Skill establishes unauthorized persistence across sessions via cron jobs, startup scripts, or state files. Session persistence allows an attacker to maintain access beyond the current interaction.

Content

Scanner excerpt · README.md (reported line 9)May include surrounding context.

md

You've done real work — repos, papers, side projects, design docs — but when it comes to describing yourself, you draw a blank. Sound familiar?

- **"I don't know what my strengths are called."** You've built ETL pipelines, but you'd never think to write "data pipeline orchestration" on your resume.
- **"I can't explain what I do."** You know you're good, but turning that into a pitch feels impossible.
- **"My work looks all over the place."** Your projects seem unrelated, but you suspect there's a thread connecting them.
- **"I don't fit one clean label."** Maybe your real edge is the combination of things you do, not any single one.

Rp1

Medium

• Category: MCP Rug Pull
• Confidence: 91% confidence

Finding

The README instructs users to run npx skills add ... without pinning an exact package version. Because npx fetches the latest published package by default, a compromised or malicious upstream release could execute unexpected code during install, creating a supply-chain risk for users following the documentation.

Content

No source excerpt is available for this finding.

Rp1

Medium

• Category: MCP Rug Pull
• Confidence: 91% confidence

Finding

This is a second unpinned npx skills add reference, carrying the same supply-chain risk as the earlier install command. Users who copy-paste it may execute a newer or attacker-controlled package version than the author intended.

Content

No source excerpt is available for this finding.

Session Persistence

Medium

• Category: Rogue Agent
• Confidence: 84% confidence

Finding

The README explicitly states the skill creates ~/.career-spotlight/ on first run, reuses it, preserves existing analyses, archives report history, and tracks staleness via hashes/timestamps. Because the skill processes potentially sensitive resumes, papers, repos, and personal branding materials, persistent storage increases privacy risk if users are not clearly warned up front or given retention controls.

Content

Scanner excerpt · README.md (reported line 141)May include surrounding context.

A six-step pipeline:

text

Init → Analyze → Position → Synthesize → Write Copy → Review

1. Init — Creates ~/.career-spotlight/ on first run. Reuses it after that.

Vague Triggers

Medium

• Category: Not specified by scanner
• Confidence: 90% confidence

Finding

The trigger description is broad enough to activate on common resume, self-introduction, or branding requests, which increases the chance the skill runs in situations where the user did not specifically intend this workflow. Because the skill also reads multiple inputs and writes persistent files, over-broad invocation expands the attack surface and can lead to unexpected data processing or side effects.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium

• Category: Not specified by scanner
• Confidence: 95% confidence

Finding

The skill instructs the agent to archive and write files in the user's home directory, but it does not include an explicit user-facing warning or consent step for those filesystem changes. This can cause unexpected persistence, overwriting, or accumulation of sensitive career materials without the user realizing the skill modifies local state.

Content

No source excerpt is available for this finding.

Session Persistence

Medium

• Category: Rogue Agent
• Confidence: 93% confidence

Finding

The skill persistently stores analyses, reports, copies, and history under ~/.career-spotlight/, creating session-spanning retention of potentially sensitive personal and professional data. Even if intended for convenience, persistent storage can expose past documents, inferred narratives, and archived outputs to later misuse, unintended reuse, or privacy leakage.

Content

Scanner excerpt · SKILL.md (reported line 64)May include surrounding context.

md

2. Validate and expand sources (auto-detect document collections in directories).
3. Ask the user to set project priorities (`highlight` or `supporting`).
4. Check existing analyses for staleness (via git hash, file mtime, or URL age).
5. Run new analyses per `guides/project-analysis-guide.md`, write to `~/.career-spotlight/analyses/`.

## Step 2 — Position

Session Persistence

Medium

• Category: Rogue Agent
• Confidence: 91% confidence

Finding

The skill directs persistent file creation in ~/.career-spotlight/copies/, which is a session-persistent side effect rather than an ephemeral transformation. In skill context this is functionally expected, but it is still dangerous because it changes the local environment and may leave sensitive career data on disk without clear consent or lifecycle controls.

Content

Scanner excerpt · guides/copywriting-guide.md (reported line 263)May include surrounding context.

md

## 7. Output

### Files to write

Write four separate files to `~/.career-spotlight/copies/`:

Missing User Warnings

Medium

• Category: Not specified by scanner
• Confidence: 94% confidence

Finding

The guide explicitly instructs the agent to write and archive files under the user's home directory, including moving existing files and overwriting outputs, without requiring explicit user confirmation at execution time. Even though this appears to support normal skill functionality, it creates a real filesystem side effect and could modify or destroy user data if triggered unexpectedly or if path handling is ever broadened.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

Medium

• Category: Not specified by scanner
• Confidence: 95% confidence

Finding

The guide instructs the agent to create persistent local storage, write a README, and maintain long-lived project data under ~/.career-spotlight/. That exceeds a narrow 'discover strengths from provided materials' role and creates a durable data store containing potentially sensitive career history, which increases privacy and scope-creep risk if the agent is over-trusted or later misused.

Content

No source excerpt is available for this finding.

Session Persistence

Medium

• Category: Rogue Agent
• Confidence: 95% confidence

Finding

This is explicit session persistence: the skill creates a dedicated directory structure and stores a continuing career profile locally. Persistent storage of analyses, copy, and history can expose sensitive professional data across sessions and widens the consequences of any later compromise or accidental access.

Content

Scanner excerpt · guides/input-collection-guide.md (reported line 13)May include surrounding context.

md

1. Tell the user: "This skill stores your career profile at `~/.career-spotlight/`. This includes per-project analyses, an aggregated report, and ready-to-use copy. The directory is local to your machine and not synced anywhere."
2. Wait for user confirmation before proceeding.
3. Create directories: `~/.career-spotlight/analyses`, `~/.career-spotlight/copies`, `~/.career-spotlight/history`.
4. Write a short README.md to `~/.career-spotlight/README.md` explaining the directory structure.
5. If directory creation fails, report the error, suggest checking permissions, and do NOT proceed.

Context-Inappropriate Capability

Medium

• Category: Not specified by scanner
• Confidence: 89% confidence

Finding

The docx-processing step invokes external commands (pandoc and a python3 one-liner) on user-supplied paths, introducing command execution and parser exposure beyond simple content reading. Even with quoted variables, this expands the attack surface through external tooling, untrusted document parsing, and environment-dependent behavior not obviously required by the skill's branding-focused purpose.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

Medium

• Category: Not specified by scanner
• Confidence: 92% confidence

Finding

The staleness-detection logic relies on shell commands for git, stat, find, sort, and head over user-influenced paths, adding repository inspection and command-execution capability beyond the core analysis task. This increases risk of unintended access to local filesystem metadata, brittle cross-platform behavior, and misuse if path handling or execution boundaries are weak elsewhere in the system.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

Medium

• Category: Not specified by scanner
• Confidence: 93% confidence

Finding

The guide directs the agent to write analyses and update metadata in a persistent profile directory, extending behavior from analysis into stateful content generation and record maintenance. This broadens the skill's authority and can silently accumulate sensitive professional information or overwrite prior data, especially if users believe the skill is only doing temporary analysis.

Content

No source excerpt is available for this finding.

Session Persistence

Medium

• Category: Rogue Agent
• Confidence: 96% confidence

Finding

Writing each analysis to a persistent per-project file creates a durable record of user-provided materials and inferred career narratives. Because these outputs may contain sensitive summaries, project details, or derived personal-branding content, retention raises privacy and data-governance concerns beyond the immediate task.

Content

Scanner excerpt · guides/input-collection-guide.md (reported line 127)May include surrounding context.

md

For each new project:

1. Read `guides/project-analysis-guide.md` and follow its methodology.
2. Write each analysis to `~/.career-spotlight/analyses/[slugified-name].md` using `templates/project-analysis.md`.

### Naming rules

Description-Behavior Mismatch

Medium

• Category: Not specified by scanner
• Confidence: 96% confidence

Finding

The guide instructs the agent to write output to a fixed path in the user's home directory even though the skill is framed as analysis/synthesis. That creates an integrity risk by modifying local files without an explicit user-approved action boundary, and it expands the skill from read/analyze into persistent side effects.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium

• Category: Not specified by scanner
• Confidence: 97% confidence

Finding

The skill directs writes to the user's home directory without a user-facing warning that local files will be created or modified. Silent file modification is dangerous because users may invoke a seemingly analytical skill without realizing it performs persistent changes on disk.

Content

No source excerpt is available for this finding.

Session Persistence

Medium

• Category: Rogue Agent
• Confidence: 90% confidence

Finding

The guide explicitly establishes session persistence by checking for an existing report, archiving it, and writing a new artifact. Persistent state is more dangerous in this skill context because the content is likely to contain sensitive personal and professional narratives, and retention is not clearly bounded or consented.

Content

Scanner excerpt · guides/narrative-synthesis-guide.md (reported line 410)May include surrounding context.

md

---

## 9. Output — Write the Report

### Pre-Write Check

Missing User Warnings

Medium

• Category: Not specified by scanner
• Confidence: 97% confidence

Finding

The archive-and-rewrite workflow replaces prior output state without an explicit warning or confirmation, which can surprise users and create unintended persistence of old versions. This is especially risky for personal career documents that may contain sensitive information or drafts the user did not intend to retain.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

Medium

• Category: Not specified by scanner
• Confidence: 88% confidence

Finding

The guide instructs the skill to scan ~/.career-spotlight/analyses/*.md and grep prior analysis content to normalize tags across projects, which expands access beyond the current user-provided artifact. This cross-project home-directory enumeration can expose unrelated sensitive data and violates least-privilege if the skill's stated purpose is analyzing a current project, not mining prior local history.

Content

No source excerpt is available for this finding.

Session Persistence

Medium

• Category: Rogue Agent
• Confidence: 94% confidence

Finding

This is a session-persistence behavior: the skill is told to immediately write analysis results to disk rather than keeping them transient. Persistent storage increases privacy and data-retention risk because project-derived summaries may contain proprietary or personal information and remain on disk beyond the user's expectation.

Content

Scanner excerpt · guides/project-analysis-guide.md (reported line 360)May include surrounding context.

md

### File Creation

After completing the five-dimension extraction for a project, immediately write the analysis to disk. Do not hold multiple analyses in context -- write each one as it is completed.

**File path:** `~/.career-spotlight/analyses/[slugified-name].md`

Description-Behavior Mismatch

Medium

• Category: Not specified by scanner
• Confidence: 92% confidence

Finding

The guide explicitly instructs the skill to persist generated analyses under the user's home directory, which exceeds a narrowly scoped read/analyze workflow and creates durable side effects. Even if intended for convenience, writing files can store sensitive project details, create unwanted artifacts, and be abused to persist data without clear user consent at execution time.

Content

No source excerpt is available for this finding.

Autonomous Decision Making

Medium

• Category: Excessive Agency
• Confidence: 85% 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 · references/industry-terms-swe.md (reported line 26)May include surrounding context.

md

| User's description | Industry term |
|---|---|
| "set up auto-deploy" | CI/CD Pipeline |
| "used containers" | Containerization (Docker, OCI) |
| "managed cloud resources" | Infrastructure as Code (IaC) (Terraform, Pulumi) |
| "set up monitoring/alerts" | Observability Stack (Logging, Metrics, Tracing) |

Missing User Warnings

Low

• Category: Not specified by scanner
• Confidence: 84% confidence

Finding

This markdown file documents that the skill creates ~/.career-spotlight/, stores per-project analyses, generated copy, and archived history, which affects user data on disk. Although the behavior is described, it is presented as part of the workflow rather than as an explicit warning or caution about persistent storage and retained history.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

Low

• Category: Not specified by scanner
• Confidence: 83% confidence

Finding

The manifest description centers on surfacing hidden strengths, industry buzzwords, and career narratives for resumes and branding. Section 8 adds a capability to identify missing capabilities and positioning risks, which is adjacent but semantically broader than the stated discovery-of-strengths purpose.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.