creating-a-changeset

Pass

Audited by Gen Agent Trust Hub on Aug 3, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: The skill ingests data from repository files such as AGENTS.md and CONTEXT.md which presents an attack surface for indirect prompt injection. The agent uses this content to determine release notes and version bumps.\n
  • Ingestion points: Reads AGENTS.md, CONTEXT.md, package.json, and the output of git commands.\n
  • Boundary markers: None identified; the instructions do not include markers to isolate ingested content from agent logic.\n
  • Capability inventory: Writing markdown files to the local .changeset/ directory and executing local CLI status checks via package managers.\n
  • Sanitization: No filtering or escaping is performed on the ingested repository content before it is processed or written.\n- [COMMAND_EXECUTION]: The skill performs shell executions to assess the repository state and validate changesets. It uses commands such as git status, git diff, and package manager invocations like pnpm exec changeset status or bun changeset status. These are restricted to local assessment and validation of the changesets workflow.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 3, 2026, 10:36 PM
Security Audit — agent-trust-hub — creating-a-changeset