The Agent Skills Directory

[SAFE]: The skill contains instructional content and shell scripts designed for local Git repository management. The commands utilize common utilities such as git, jq, grep, sort, and uniq to analyze logs and verify file integrity. These operations are restricted to the local development environment and perform standard repository maintenance tasks.
[INDIRECT_PROMPT_INJECTION]: The skill defines workflows that involve reading data from the repository environment which could technically be influenced by other contributors.
Ingestion points: Git commit history via git log, registry files like manifest.json, and test files.
Boundary markers: No explicit delimiters or safety warnings for embedded instructions are present in the scripts, which is typical for developer utility snippets.
Capability inventory: Capabilities are limited to standard file system reads, local shell utilities, and build tools (make). There are no network exfiltration patterns, persistent mechanisms, or privilege escalation commands.
Sanitization: The skill uses structured processors like jq to parse JSON data, which naturally restricts the interpretation of data fields.

concurrent-branches