The Agent Skills Directory

[COMMAND_EXECUTION]: The skill provides instructions for using the codex CLI tool to perform software engineering tasks. This includes the use of sandboxing flags such as --sandbox danger-full-access and --sandbox workspace-write which allow the tool to interact with the file system and network.
[EXTERNAL_DOWNLOADS]: The skill is configured to silently check for updates by fetching a version file from the vendor's repository at https://raw.githubusercontent.com/abpai/skills/main/versions.json. This is used to notify the user of available updates.
[PROMPT_INJECTION]: The skill has an attack surface for indirect prompt injection because it summarizes and follows instructions based on the output of the codex tool.
Ingestion points: Results and summaries produced by codex exec and codex resume commands in SKILL.md.
Boundary markers: No delimiters or specific instructions are provided to the agent to ignore potential instructions embedded in the tool's output.
Capability inventory: The agent can execute shell commands and modify code files via the codex tool.
Sanitization: The skill does not implement validation or filtering of the content returned by the CLI tool.

codex