The Agent Skills Directory

[SAFE]: The skill follows security best practices by enforcing a read-only boundary and explicitly routing security-sensitive audits to dedicated skills. No malicious code or exfiltration patterns were identified.
[COMMAND_EXECUTION]: The skill instructions include the use of common, non-destructive git commands (e.g., "git diff", "git rev-parse") to gather repository context. These commands are standard for code analysis and do not pose a security risk in this context.
[PROMPT_INJECTION]: The skill analyzes external code, which is a potential surface for indirect prompt injection (Category 8). However, the risk is considered safe given the skill's limited read-only scope and focus on reporting refactoring opportunities rather than execution.
Ingestion points: Processes project manifests (package.json, etc.), source files, and git diffs as described in Step 1 and Step 2.
Boundary markers: Encourages a "read-only boundary" check but lacks explicit data delimiters for ingested code.
Capability inventory: Capabilities are limited to file reading and informational git commands; it specifically prohibits code modification.
Sanitization: No explicit sanitization or filtering of ingested code content is instructed.

code-slimming