The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes several local shell commands and scripts to perform its tasks.
Execution of scripts/audit-repo.sh to generate audit reports.
Use of the GitHub CLI (gh) for cloning, creating pull requests, checking status, and modifying repository settings via API.
Execution of git commands for branching, committing, and pushing changes.
[REMOTE_CODE_EXECUTION]: The skill performs local execution of build and linting tools on code cloned from external repositories.
Specifically, it runs cargo fmt, clippy, and cargo test on Rust repositories.
Running cargo test is a known risk as it can execute arbitrary code defined in build.rs or test suites within the cloned repository.
[EXTERNAL_DOWNLOADS]: The skill references and fetches configuration templates from external GitHub repositories.
It uses templates for CODEOWNERS, dependabot, and other governance files from the third-party repository cislunarspace/CODE-core.
It also references templates from ouyangjiahong26/altgo, which belongs to the skill author.
[INDIRECT_PROMPT_INJECTION]: The skill processes data from untrusted sources, creating a potential injection surface.
Ingestion points: Cloned content from external GitHub repositories and output from the audit-repo.sh script (SKILL.md).
Boundary markers: None identified in the instructions for separating untrusted data from agent instructions.
Capability inventory: Includes file system writes, network operations via gh API, and command execution (SKILL.md).
Sanitization: No explicit sanitization or validation of the content being merged from external templates or repositories is mentioned.

github-gov-fix