backblaze-coder

Pass

Audited by Gen Agent Trust Hub on Oct 9, 2026

Risk Level: SAFEDATA_EXFILTRATIONPRIVILEGE_ESCALATIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
  • [DATA_EXFILTRATION]: The skill references and provides instructions for managing sensitive configuration files that store credentials, including ~/.aws/credentials, ~/.config/rclone/rclone.conf, and ~/.b2_account_info.
  • [PRIVILEGE_ESCALATION]: The installation documentation in references/b2-cli-setup.md encourages the use of sudo to install the Backblaze B2 CLI on Linux systems (sudo apt install backblaze-b2).
  • [EXTERNAL_DOWNLOADS]: The skill facilitates the installation of external software packages, specifically the b2 tool via pip and b2-tools via Homebrew.
  • [INDIRECT_PROMPT_INJECTION]: The skill provides an attack surface for indirect prompt injection by facilitating the download and synchronization of content from remote cloud storage buckets.
  • Ingestion points: Commands like b2 download-file-by-name and b2 sync (from B2 to local) in SKILL.md bring external, potentially untrusted data into the local environment.
  • Boundary markers: There are no explicit instructions or delimiters used to warn the agent about ignoring potential instructions embedded in downloaded files.
  • Capability inventory: The skill uses the Bash tool and performs file system write operations, which could be leveraged if the agent obeys instructions found in synchronized data.
  • Sanitization: The skill does not describe any sanitization or validation steps for the content fetched from remote buckets.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 9, 2026, 12:28 AM