acreadiness-generate-instructions
Pass
Audited by Gen Agent Trust Hub on Sep 17, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONREMOTE_CODE_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill fetches the AgentRC tool from the Microsoft GitHub repository using npx.
- [COMMAND_EXECUTION]: Executes shell commands to invoke the instruction generator and manage directory structures for instruction files.
- [REMOTE_CODE_EXECUTION]: Executes code retrieved from the Microsoft GitHub organization via the npx package runner.
- [INDIRECT_PROMPT_INJECTION]: The skill ingests repository source code and local configuration files to generate tailored agent instructions, creating a potential path for untrusted data to influence agent behavior.
- Ingestion points: Reads agentrc.config.json and the repository's source code files during the generation process.
- Boundary markers: Absent; the skill does not wrap generated instructions in specific delimiters or include warnings to ignore embedded content.
- Capability inventory: Command execution via npx, file writing, and directory management within the .github and .agents paths.
- Sanitization: Absent; content from the repository and config file is processed and interpolated into the output instructions without explicit filtering.
Audit Metadata