The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill represents a potential attack surface because it ingests and modifies code files which could contain malicious instructions designed to influence the agent's behavior.
Ingestion points: The agent uses Read, Glob, and Grep tools to process code files as described in SKILL.md.
Boundary markers: The skill includes a "Scope guardrail" explicitly directing the agent to never modify executable code, identifiers, imports, or logic, effectively creating a logical boundary for its actions.
Capability inventory: The skill uses Write and Edit tools to modify the contents of files.
Sanitization: The skill provides a "Code Comment Checklist" to ensure modifications are limited to the intended scope and follow safety protocols.
[PROMPT_INJECTION]: No evidence of instructions designed to bypass safety filters or override system behavior was found. The skill focuses on style and hygiene for code documentation.
[DATA_EXFILTRATION]: No network operations, credential access, or sensitive file path access were detected.
[COMMAND_EXECUTION]: The skill uses file manipulation tools but does not execute shell commands or interact with the operating system beyond reading and writing files as directed by its primary purpose.
[OBFUSCATION]: No encoded content, hidden characters, or homoglyphs were found in the skill documentation.

antislop-code