add-new-model
Pass
Audited by Gen Agent Trust Hub on Oct 14, 2026
Risk Level: SAFECREDENTIALS_UNSAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONDYNAMIC_EXECUTION
Full Analysis
- [CREDENTIALS_UNSAFE]: The skill directs the agent to read sensitive provider API keys from a repository-root
.envfile using thesource .envcommand to authenticate verification calls. - [EXTERNAL_DOWNLOADS]: Network requests are performed using
curl,awsCLI, andghAPI to fetch model information and pricing data from trusted organizations including OpenAI, Anthropic, Google, and the project's own pricing repository. - [COMMAND_EXECUTION]: Shell commands for searching and verification (e.g.,
grep,rg,git log,uv run) are constructed using user-provided model identifiers and provider names. - [DYNAMIC_EXECUTION]: The instructions require the agent to dynamically generate and execute local Python scripts (e.g.,
local-notes/probe_<model>.py) to compare model capabilities and API behaviors at runtime.
Audit Metadata