release
Pass
Audited by Gen Agent Trust Hub on Apr 6, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes multiple local shell scripts (e.g.,
scripts/release.sh,scripts/install-hooks.sh) to perform versioning and release operations. - [COMMAND_EXECUTION]: It installs a local git pre-push hook by creating a symlink in the
.git/hooks/directory and applyingchmod +xto the source script. - [COMMAND_EXECUTION]: Interacts with remote repositories and CI services through network-enabled commands such as
git pushand the GitHub CLI (gh). - [PROMPT_INJECTION]: The skill contains a vulnerability surface for indirect prompt injection through data ingestion. 1. Ingestion points:
SKILL.mdprocess readsCHANGELOG.mdand commit history viascripts/release-context.sh. 2. Boundary markers: Absent: No delimiters or ignore instructions are used when processing external text. 3. Capability inventory: Subprocess calls (git, gh) and file-write operations. 4. Sanitization: Absent: No sanitization of the changelog or commit message content is performed.
Audit Metadata