create-lib-docs
Pass
Audited by Gen Agent Trust Hub on Nov 9, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONDATA_EXFILTRATIONDYNAMIC_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill runs the library's CLI with the
--helpflag and executes code snippets to verify their output. This behavior is essential for the skill's purpose of generating verified documentation. - [INDIRECT_PROMPT_INJECTION]: The skill reads and processes untrusted data from the library's source code, tests, and CLI outputs to build a documentation inventory.
- Ingestion points: Library source files,
package.json,pyproject.toml, and CLI help output. - Boundary markers: The skill instructs the agent to capture real output but does not specify explicit delimiters for untrusted content.
- Capability inventory: Includes file reading/writing and local shell command execution.
- Sanitization: Documentation verification scripts check for link integrity and formatting but do not sanitize for prompt injection.
- [DATA_EXFILTRATION]: The skill accesses project configuration files and source code to identify public APIs and configuration keys. All operations are local, and no network exfiltration patterns were detected.
- [DYNAMIC_EXECUTION]: Local Python scripts are used to lint documentation and generate
llms.txtandllms-full.txtfiles. These scripts rely on the standard library and perform deterministic file processing.
Audit Metadata