Data Pipeline Toolkit
Security checks across malware telemetry and agentic risk
Overview
The skill describes a broad data-pipeline tool but ships no pipeline code while instructing agents to run a local script and use database/cloud credentials, so it needs review before use.
Treat this as incomplete until the actual pipeline.sh implementation and install source are available for review. If you still use it, run commands manually, verify the script path, use test data first, and provide only narrowly scoped credentials.
VirusTotal
VirusTotal findings are pending for this skill version.
Risk analysis
Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.
A user or agent could end up running unreviewed or unintended local code while believing it is part of the installed skill.
The central workflow depends on a local pipeline.sh helper, but the provided manifest contains only SKILL.md and _meta.json with no install spec or code file for that helper. Running this command would execute whatever local script exists rather than reviewed package code.
./pipeline.sh create my-pipeline
Do not run the quick-start commands unless you know where pipeline.sh comes from. The skill should include the helper script or a pinned, auditable install source and declare its runtime requirements.
ETL use may require credentials that can read or modify important databases, APIs, or cloud storage locations.
The example expects a database connection string, and the skill lists databases, APIs, S3, Google Cloud Storage, Kafka, and SQS as sources or destinations, but the registry metadata declares no required credentials or environment variables.
./pipeline.sh load my-pipeline postgres --connection $DB_URL
Use narrowly scoped credentials, confirm the exact source and destination before running a pipeline, and avoid giving production credentials to an unreviewed helper.
A pipeline could keep running on a schedule and continue accessing or moving data after the user’s immediate request is complete.
Scheduling is an expected ETL feature, but cron-based or event-driven jobs can continue running after the initial task if not explicitly reviewed and removed.
- **Scheduling** - Cron-based or event-driven
Only enable schedules intentionally, document where they are created, and provide a clear way to list, pause, and delete scheduled jobs.