opencode
Pass
Audited by Gen Agent Trust Hub on Sep 1, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill instructs the agent to install the
opencode-aipackage via NPM and a Homebrew tap (anomalyco/tap/opencode) from external sources not belonging to the trusted vendor list. - [COMMAND_EXECUTION]: The skill extensively uses the
terminaltool to execute shell commands for tool setup, git operations, and autonomous coding tasks. - [INDIRECT_PROMPT_INJECTION]: 1. Ingestion points: The skill facilitates cloning and reviewing external repositories and pull requests (
git clone,opencode pr). 2. Boundary markers: Absent; there are no instructions to use delimiters or ignore embedded commands in the processed data. 3. Capability inventory: The agent possesses shell execution (terminal) and process management (process) tools. 4. Sanitization: Absent; external code is processed by the coding agent without validation or filtering.
Audit Metadata