aube-bump
Warn
Audited by Gen Agent Trust Hub on Aug 27, 2026
Risk Level: MEDIUMREMOTE_CODE_EXECUTIONINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [REMOTE_CODE_EXECUTION]: The skill fetches source code from a third-party repository (https://github.com/jdx/aube) and executes it through
cargo check,cargo clippy, andcargo testduring the verification phase.\n- [INDIRECT_PROMPT_INJECTION]: The skill ingests external content from a remote repository and delegates the resolution of merge conflicts to secondary AI agents. This creates a surface where malicious instructions embedded in the upstream code (such as in comments, tests, or documentation) could influence the behavior of the resolving agents.\n - Ingestion points: External source code is fetched from
aube-upstreaminSKILL.md.\n - Boundary markers: None identified; the skill does not specify markers to isolate external content from instructions given to resolver agents.\n
- Capability inventory: The workflow includes powerful capabilities such as
cargo test(arbitrary code execution),git merge, andrsyncfor file system modification.\n - Sanitization: There is no evidence of automated sanitization or filtering of the incoming upstream delta before it is processed by the agent.
Audit Metadata