create-specs

Pass

Audited by Gen Agent Trust Hub on Mar 29, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: Downloads shell scripts from the official GitHub spec-kit repository (github.com/github/spec-kit) to enable automation slash commands.
  • [COMMAND_EXECUTION]: Performs standard file and directory management tasks using mkdir, chmod, cat, and jq for project initialization and configuration parsing.
  • [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface by processing content from docs/reverse-engineering/functional-specification.md. Malicious instructions embedded in these documents could potentially influence the agent during the transformation process.
  • Ingestion points: Reads the project's functional specification file to generate feature specs (SKILL.md).
  • Boundary markers: No explicit boundary markers or isolation instructions are present to delimit untrusted documentation content.
  • Capability inventory: includes file system writes (mkdir), network operations (curl), and shell script execution (SKILL.md).
  • Sanitization: Does not describe specific validation or filtering steps for the ingested documentation content.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 29, 2026, 05:28 PM
Security Audit — agent-trust-hub — create-specs