The Agent Skills Directory

[DYNAMIC_EXECUTION]: The skill uses eval and source commands in the preamble and throughout the workflow to execute output from and load environment variables from local binaries located in the vendor's directory at ~/.claude/skills/gstack/bin/.
[COMMAND_EXECUTION]: The skill makes extensive use of the Bash tool to perform file system operations (mkdir, touch, rm), git repository analysis (git diff, git log), and interactions with GitHub (gh) and GitLab (glab) command-line interfaces.
[INDIRECT_PROMPT_INJECTION]: The skill possesses a vulnerability surface for indirect prompt injection as it ingests and processes untrusted data from git diff and git log to automate documentation updates.
Ingestion points: Data enters the context via git diff and git log in Steps 1 and 1.5, as well as existing project documentation.
Boundary markers: The instructions do not define explicit delimiters or 'ignore' warnings for the data extracted from the diffs.
Capability inventory: The skill has access to Bash, Write, and Edit tools, allowing it to modify the file system and execute shell commands based on its analysis.
Sanitization: There is no evidence of escaping or filtering of the external content retrieved from the git history before it is used to influence documentation updates.
[PERSISTENCE]: The skill can persistently alter agent behavior across sessions by modifying the project's CLAUDE.md file to add specific 'Skill routing' instructions. It also maintains state through the creation of marker files and analytics logs in the ~/.gstack/ directory.
[EXTERNAL_DOWNLOADS]: The skill triggers an update check via the gstack-update-check binary, which involves network communication to verify the local version against a remote registry.

document-release