write-incident-report

Pass

Audited by Gen Agent Trust Hub on Aug 17, 2026

Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill utilizes common CLI tools (ls, sed, rg, git diff) to perform repository checks and validation of the generated Markdown files. These operations are restricted to the local file system and are used to ensure compliance with the repository's documentation standards.
  • [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to read existing incident reports in docs/incidents/ and the README.md to determine next IDs and maintain consistency.
  • Ingestion points: Reads files within the docs/incidents/ directory.
  • Boundary markers: None identified; the agent is simply instructed to "inspect" the files.
  • Capability inventory: File creation and writing in docs/incidents/, and shell execution of ls, sed, rg, and git.
  • Sanitization: No specific sanitization or filtering is performed on the content read from existing reports. While this presents a minor theoretical risk if the repository contains malicious content, the highly structured template required for the output significantly restricts the potential for successful exploitation.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 17, 2026, 10:58 PM
Security Audit — agent-trust-hub — write-incident-report