The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to execute a shell command (date -u) to generate formatted UTC timestamps for its operation log entries.
[DATA_EXFILTRATION]: The evidence action enables the agent to read contents from arbitrary local file paths provided by the user and transmit that data to the Ghostwriter logging service via the create_oplog_entry tool.
[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from local files and passes it to subsequent tools, creating a vulnerability surface for instruction injection.
Ingestion points: Local file contents read during the execution of the evidence action (SKILL.md).
Boundary markers: Absent; the instructions do not specify delimiters or protective wrappers for the ingested file content.
Capability inventory: Local file read access, execution of the date command, and access to the create_oplog_entry MCP tool.
Sanitization: The skill does not provide instructions for sanitizing or validating file content before it is processed by the agent or the logging tool.

ghostwriter-oplog