The Agent Skills Directory

[COMMAND_EXECUTION]: The skill automatically executes a shell command to synchronize repository data.
Evidence: git -C <ceos_root> pull --ff-only --quiet 2>/dev/null is invoked to ensure local data is up to date with teammates.
[EXTERNAL_DOWNLOADS]: The skill performs network operations to pull updates from a remote git repository.
Evidence: The git pull operation fetches data from a remote source, though it is intended for synchronization within the CEOS framework.
[PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection because it processes untrusted content from markdown files to drive its conversation logic.
Ingestion points: The skill reads data/vision.md, data/accountability.md, data/rocks/, and data/people/ to extract core values, seats, and performance ratings.
Boundary markers: Absent. The skill does not use specific delimiters or instructions to ignore potential commands embedded in the source files.
Capability inventory: The skill has the ability to Write to the file system (creating conversation logs) and execute git commands via a subprocess.
Sanitization: Absent. There is no evidence of validation or filtering for the content extracted from the markdown files before it is used in the agent's prompts.

ceos-quarterly