The Agent Skills Directory

[COMMAND_EXECUTION]: The skill's operational logic depends on the execution of local system commands via Python's subprocess module. (1) scripts/check_dependencies.py executes ruby to verify the availability of the Psych YAML library. (2) scripts/insert_frontmatter.py executes git to extract author and committer timestamps from repository history. (3) scripts/okf_yaml.py executes an embedded Ruby script via subprocess.run for robust YAML parsing and duplicate key detection. (4) scripts/test_okf.py invokes git to set up test fixtures and runs the other skill scripts.
[PROMPT_INJECTION]: The skill presents a surface for indirect prompt injection as it processes untrusted Markdown content to suggest metadata. (1) Ingestion points: Markdown file content is read from the local filesystem in scripts/insert_frontmatter.py and scripts/validate_okf.py. (2) Boundary markers: While references/markup-agent-prompt.md provides guidance, the skill does not use explicit data delimiters when processing file content for metadata derivation. (3) Capability inventory: The skill can write to files and execute local commands (git, ruby, python). (4) Sanitization: The skill employs YAML.safe_load patterns and performs explicit validation on metadata fields.

x9-okf-adapt