The Agent Skills Directory

[PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection (Category 8) because it processes untrusted data from the local filesystem.\n
Ingestion points: Reads feature context from .chalk/docs/product/, .chalk/docs/engineering/, and implementation source code.\n
Boundary markers: No instructions are provided to distinguish skill commands from audited content or to ignore embedded instructions.\n
Capability inventory: The skill has access to Read, Glob, Grep, and Write tools to interact with the local filesystem.\n
Sanitization: No validation or sanitization is performed on ingested content before it is processed by the agent.- [SAFE]: No direct prompt injection, obfuscation, or metadata poisoning was detected.- [SAFE]: No network operations, hardcoded credentials, or data exfiltration patterns were found.- [SAFE]: No external downloads, package installations, or remote code execution patterns are present.- [SAFE]: The skill uses standard file system tools for its intended purpose and does not attempt privilege escalation or persistence.

audit-accessibility