read-build-logs-bottom-up

Pass

Audited by Gen Agent Trust Hub on Aug 30, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process the contents of a build log file, creating a surface for indirect prompt injection attacks. 1. Ingestion points: The skill reads data from a file defined by the $LOG variable (e.g., build.log) using grep, sed, tail, and head in SKILL.md. 2. Boundary markers: There are no explicit boundary markers or instructions to the agent to disregard commands or instructions embedded within the build log output. 3. Capability inventory: The skill uses standard shell utilities (ls, wc, grep, tail, head, sed) to extract and display information from the filesystem. 4. Sanitization: No sanitization is performed on the log content to prevent instructions within the log from being interpreted by the agent.
  • [COMMAND_EXECUTION]: The skill uses standard shell utilities (grep, sed, ls, wc, tail, head) to inspect file metadata and content. These are used safely in the context of log analysis.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 30, 2026, 07:03 AM
Security Audit — agent-trust-hub — read-build-logs-bottom-up