The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes a wide range of shell commands to manage the release lifecycle, including interaction with the GitHub CLI (gh), build tools, and package managers.
[DATA_EXFILTRATION]: The instructions direct the agent to access the user's shell profile (~/.zprofile) to retrieve sensitive NPM_TOKEN and PYPI_TOKEN credentials. Access to shell configuration files that store secrets is a sensitive operation.
[REMOTE_CODE_EXECUTION]: The skill relies on and executes repository-local scripts, specifically scripts/publish-release.sh and scripts/prepare-app-bundle.mjs, which contain the core logic for building and publishing the software.
[PROMPT_INJECTION]: The skill is subject to indirect prompt injection risks because it reads and processes data from project files like CHANGELOG.md to automate the generation of release notes and commit messages.
Ingestion points: Reads content from CHANGELOG.md, package.json, and pyproject.toml (SKILL.md).
Boundary markers: No delimiters or safety instructions are defined to prevent the agent from following instructions embedded in the ingested files.
Capability inventory: The skill possesses extensive capabilities, including subprocess execution (bash, npm, python), file system writes (version updates), and network operations (gh, npm, twine).
Sanitization: No validation or sanitization of the content read from external project files is specified.

publish-orion-release