gpc-release-flow

Pass

Audited by Gen Agent Trust Hub on May 18, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [PROMPT_INJECTION]: The skill features an indirect prompt injection surface in commands that ingest untrusted external data. For example, gpc changelog generate processes git commit history, and various release commands read content from local release notes files. If these sources contain adversarial instructions, they could potentially influence the agent's logic during processing. (Ingestion points: Git logs in SKILL.md, release notes files in references/release-notes.md; Boundary markers: Absent; Capability inventory: Shell command execution via gpc CLI in SKILL.md; Sanitization: Absent).
  • [COMMAND_EXECUTION]: The skill utilizes the gpc CLI for all operations. A detection script (scripts/detect_gpc.mjs) is included which executes shell commands to identify the CLI version and authentication state. These commands are static and do not interpolate user input.
  • [EXTERNAL_DOWNLOADS]: Instructions within the skill and its references guide users to install the @gpc-cli/cli package from the public NPM registry. This is a standard dependency for the skill's functionality.
Audit Metadata
Risk Level
SAFE
Analyzed
May 18, 2026, 09:09 AM
Security Audit — agent-trust-hub — gpc-release-flow