The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes several command-line tools including git, gh (GitHub CLI), glab (GitLab CLI), and jq to perform its setup tasks.
It executes git remote -v and reads .git/config to identify the repository's origin.
It uses gh or glab templates to interact with remote issue trackers for project management.
It employs a shell script involving jq to safely merge security restrictions into the .claude/settings.json file, which is a standard procedure for configuring project-specific agent permissions.
[INDIRECT_PROMPT_INJECTION]: The skill has a surface for indirect prompt injection as it ingests data from local project files (such as CLAUDE.md and CONTEXT.md) and external issue trackers.
Ingestion points: Reads repository metadata, local documentation files, and issue/pull request content from GitHub or GitLab.
Boundary markers: The skill explicitly requires a human-in-the-loop workflow, instructing the agent to present findings and draft content to the user for confirmation and editing before any write operations occur.
Capability inventory: The skill can read/write local files and execute specific CLI tools for source control and issue management.
Sanitization: Relies on user review of proposed changes to configuration files and documentation before execution.

setup-skills