The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes multiple local shell commands using tools like Git, Bun, and the GitHub CLI (gh) to manage the release state and verify artifacts. This includes running internal project scripts such as generate-changelog.ts and print-release-notes.ts to process repository data.
[REMOTE_CODE_EXECUTION]: Triggers a GitHub Actions workflow (publish.yml) on the repository code-yeongyu/oh-my-openagent. This constitutes remote execution of build and deployment logic on a trusted platform managed by the vendor.
[DATA_EXFILTRATION]: Performs network operations to send release summaries to a Discord channel via the agent-discordbot tool. This is an intended communication feature for notifying stakeholders of new releases and uses a specific bot ID (1486173823354146917).
[INDIRECT_PROMPT_INJECTION]: The skill analyzes and summarizes repository data (commit messages and changelogs) to generate user-facing release notes.
Ingestion points: Data enters the context from the outputs of project-local scripts in Step 5.
Boundary markers: No specific delimiters or "ignore instructions" warnings are used for the interpolated changelog data during the summary drafting phase.
Capability inventory: The skill has access to file system writes and shell command execution through various CLI tools.
Sanitization: No explicit sanitization or filtering of the processed commit messages is described before interpolation into the drafting prompt, though the process is gated by a procedural review step.

publish