snapshot-writer
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFE
Full Analysis
- [Data Exposure & Exfiltration] (SAFE): The skill only interacts with local project files (CSV, YAML, JSONL) within the workspace and explicitly declares no network access. No exfiltration risk is present.- [Indirect Prompt Injection] (LOW): The skill ingests data from external files (
outline/outline.yml,papers/core_set.csv). While these are potential ingestion points for untrusted content, the skill lacks dangerous capabilities (network access, shell execution, or sensitive file reads) required for an exploit to be impactful. - Ingestion points: Reads
outline/outline.yml,papers/core_set.csv, andpapers/papers_dedup.jsonl. - Boundary markers: Absent; the role prompts do not explicitly define delimiters for these inputs.
- Capability inventory: Limited to writing a markdown file to
output/SNAPSHOT.md. - Sanitization: None specified.- [Remote Code Execution] (SAFE): No remote scripts or external dependencies are referenced or executed.- [Command Execution] (SAFE): No shell commands or subprocess calls are present in the skill definition.
Audit Metadata