The Agent Skills Directory

[COMMAND_EXECUTION]: The skill makes extensive use of shell commands including git and gh CLI to manage branches, commits, pulls, and pull requests. It also executes local scripts such as scripts/teardown.sh and scripts/worktree-remove.sh during the --close workflow.
[EXTERNAL_DOWNLOADS]: Recommends the execution of npx -y github:timschoch/skilly add babysit-pr if a dependency is missing. This command downloads and executes code from a GitHub repository associated with the skill's author.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from Git diffs, PR titles, and PR bodies to generate Slack summaries and PR descriptions.
Ingestion points: Data is ingested via gh pr view, gh pr list, and git diff commands.
Boundary markers: No explicit boundary markers or instruction-ignore warnings are used when interpolating this data into prompts for summarization.
Capability inventory: The skill has the capability to execute shell commands, read/write files (including .env), and post messages to Slack via MCP tools.
Sanitization: There is no explicit sanitization or filtering of the ingested content. However, the skill implements a mandatory human-in-the-loop step by showing a preview of the Slack message before posting.
[DATA_EXFILTRATION]: The skill manages Slack notifications and interacts with environment variables. It may append a user's Slack ID to the .env file to persist configuration. It sends PR details (titles, URLs, summaries) to a Slack channel (#dev).
[DYNAMIC_EXECUTION]: The skill utilizes dynamic script execution by sourcing shell scripts (e.g., . .agents/hooks/lib/branch-name.sh) and running a Node.js script from a local directory (node .agents/skills/verify/scripts/default-branch.mjs) to determine repository configuration.

git-shortcuts