100Critical
This skill performs unauthorized schema discovery and executes dangerous SQL injection patterns to exfiltrate sensitive production data under the guise of generating documentation.
Skill Info
Nameanthropics/knowledge-work-plugins/data-context-extractor
Registrygithub
Version55447a28e6c7
PURLpkg:github/anthropics/knowledge-work-plugins@55447a28e6c7?skill=data-context-extractor
Stars21,418
Installs1,900
Source
Install
This skill has been flagged as potentially malicious. Review the findings below before installing.
GitHub
1,900 installs
Skills.shDocs
npx skills add https://github.com/anthropics/knowledge-work-pluginsAssessments (3)
AI Agent Traps ↗SQL injection — database system-object or file-access abuse (information_schema/xp_cmdshell/INTO OUTFILE/LOAD_FILE) detectedcriticalsql injectionAML.T0050AML.T0053LLM05:2025
SQL injection — database system-object or file-access abuse (information_schema/xp_cmdshell/INTO OUTFILE/LOAD_FILE) detected (seen 2 times in this file at lines 59, 62)
100% confidenceline 59
INFORMATION_SCHEMA.SCHEMATA
Data Exfiltration via Schema Discoveryhighdata exfiltrationAML.T0025AML.T0024LLM02:2025
The skill instructs the agent to perform broad schema discovery and extract 'tribal knowledge' (business logic, metrics, and table relationships) from a production data warehouse. This creates a high risk of exfiltrating sensitive business intelligence and proprietary data structures into a portable, user-controlled zip file.
80% confidence
Step 2: Explore the schema... List available datasets/schemas... Identify the most important tables... Pull schema details for those key tables
Capability Inflation / Over-Privileged Data Accessmediumdescription mismatchAML.T0011.000LLM06:2025LLM09:2025
The skill claims to be a 'meta-skill' for generating documentation, but it mandates active interaction with production databases (querying schemas, identifying tables). This is a significant capability inflation beyond simple text generation.
85% confidence
Use ~~data warehouse tools (query and schema) to connect.
Description Mismatchhighdescription mismatchAML.T0011.000LLM06:2025LLM09:2025
The skill encourages the execution of arbitrary schema discovery queries against a user's production data warehouse, which poses a significant risk of data exfiltration or unauthorized metadata harvesting if the user provides broad database credentials.
85% confidence
The skill explicitly instructs the AI to use '~~data warehouse' tools to query INFORMATION_SCHEMA and system tables (e.g., 'SELECT schema_name FROM INFORMATION_SCHEMA.SCHEMATA') to map the user's entire data environment.
Social Engineering for Credential/Access Abusemediumsocial engineeringAML.T0052.000LLM09:2025
The skill uses a 'Bootstrap Mode' that encourages the agent to probe for database connection details and warehouse types, potentially tricking the user into providing credentials or granting the agent excessive permissions to sensitive production databases.
70% confidence
Step 1: Identify the database type... Use ~~data warehouse tools (query and schema) to connect.
Missing licenseinfopolicy violation
Skill does not specify a license field. Specifying a license helps users understand usage terms.
100% confidence
Missing or insufficient descriptioninfopolicy violation
Skill description is empty or too short. A clear description helps users evaluate the skill's purpose.
100% confidence
Badge
Markdown
[](https://mondoo.com/ai-agent-security/skills/github/anthropics/knowledge-work-plugins/data-context-extractor)HTML
<a href="https://mondoo.com/ai-agent-security/skills/github/anthropics/knowledge-work-plugins/data-context-extractor"><img src="https://mondoo.com/ai-agent-security/api/badge/github/anthropics/knowledge-work-plugins/data-context-extractor.svg" alt="Mondoo Skill Check" /></a>Image URL
https://mondoo.com/ai-agent-security/api/badge/github/anthropics/knowledge-work-plugins/data-context-extractor.svgSecure your AI agents
Skills can read files, run commands, and access credentials. Mondoo helps organizations manage the security risks of AI agent skills across their entire fleet.
- Continuous scanning of skills across all registries
- Policy enforcement before skills reach your agents
- Integration with your existing security workflow