The Agent Skills Directory

[COMMAND_EXECUTION] (MEDIUM): The skill's primary purpose is to execute shell commands specified in a .asc/workflow.json file.
Evidence: The documentation describes the run step executing commands via bash -o pipefail -c.
Risk: This allows for arbitrary command execution on the host machine based on repository content.
[PROMPT_INJECTION] (LOW): The skill is vulnerable to indirect prompt injection by processing untrusted data from repository files into shell execution strings.
Ingestion points: .asc/workflow.json and CLI parameters.
Boundary markers: Absent; no delimiters or warnings are used to separate instructions from untrusted data.
Capability inventory: Full shell command execution via subprocess calls.
Sanitization: Absent; the skill relies on standard shell variable expansion without input validation or escaping.

asc-workflow