The Agent Skills Directory

[SAFE]: No malicious patterns, obfuscation, or unauthorized data access were detected in the instructions. The skill accurately reflects its stated purpose of coordinating other tools.
[NO_CODE]: This skill consists entirely of markdown instructions and metadata, with no embedded scripts or executable commands. It relies on the agent to invoke other skills in a specific sequence.
[INDIRECT_PROMPT_INJECTION]: The orchestration logic relies on data ingested from previous phases (discovery and audit) to inform transitions. While this defines an attack surface where malicious data in a project's codebase could influence the agent's implementation path, the risk is inherent to the task of programmatic SEO and code analysis.
Ingestion points: User confirmation steps and output reports from pseo-discovery and pseo-audit.
Boundary markers: Not defined in this orchestration layer.
Capability inventory: Bash, Write, Edit, Read, Glob, Grep tools allowed in frontmatter.
Sanitization: Not performed within this orchestration file; assumes individual skills handle their own data validation.

pseo-orchestrate