autor3search-typescript
Pass
Audited by Gen Agent Trust Hub on Oct 13, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill installs the
@autor3search/typescriptpackage from the npm registry to facilitate benchmarking and performance measurement. - [COMMAND_EXECUTION]: The skill executes multiple shell commands, including
git status,npm install,git commit, and severalnpx autor3search-typescriptsubcommands (init,doctor,baseline,eval,report). These are used to manage the repository state and run the optimization loop. - [INDIRECT_PROMPT_INJECTION]: The skill is instructed to read and follow the contents of
program.mdas its primary instruction set for the optimization process. - Ingestion points: Reads the full content of
program.mdfrom the local repository. - Boundary markers: None identified; the skill is explicitly told to treat the file as verbatim instructions and follow it exactly.
- Capability inventory: The skill has the capability to execute shell commands, read/write files, and commit changes via Git.
- Sanitization: No validation or sanitization is performed on the instructions retrieved from the external file.
Audit Metadata