The Agent Skills Directory

[NO_CODE]: The skill package consists entirely of Markdown documentation and configuration files. None of the Python scripts referenced throughout the documentation (e.g., scripts/yao.py, scripts/governance_check.py, scripts/telemetry_native_host.py) are included in the provided files.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process untrusted external data—such as transcripts, workflows, and notes—to create or refactor agent skills. This ingestion of untrusted content into a system with high-privilege capabilities (as described in the documentation) creates a vulnerability surface for indirect prompt injection. Ingestion points: Untrusted data enters the agent context via the workflows, transcripts, and notes processed by the meta-skill as described in SKILL.md. Boundary markers: The documentation (e.g., in references/user-memory-policy.md) advocates for explicit source paths and redaction, though these markers are guidance for the user/developer rather than enforced technical boundaries within the provided files. Capability inventory: The documentation in references/trust-security-method.md and references/autonomous-adaptation.md describes capabilities including file system writes, subprocess execution, and network access. Sanitization: The methodology recommends manual review and redaction of sensitive information, but no automated sanitization logic is present in the provided files.

yao-meta-skill