The Agent Skills Directory

[DYNAMIC_EXECUTION]: The script scripts/qa-gather.sh uses the find command to locate executable Python scripts (jira-qa-gather.py or qa-gather.py) within the user's home directory and plugin cache paths at runtime. The script then executes the first matching file using uv run. This pattern of dynamic loading from computed paths is a security concern as it could be exploited to execute unintended code if the search paths are compromised or if a malicious file is placed in a high-priority search location.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze untrusted data from an external ticketing system (Jira). Ingestion points: The scripts/qa-gather.sh script (via the jira-communication companion skill) fetches issue descriptions, comments, and linked ticket data. Boundary markers: The instructions and reference files do not provide delimiters or instructions for the agent to ignore potentially malicious content embedded in the ingested ticket data. Capability inventory: The skill has access to the Bash, Read, Write, and Edit tools as defined in the skill metadata. Sanitization: No evidence of input sanitization or validation was found for the data fetched from the external ticketing system before it enters the agent's context.
[COMMAND_EXECUTION]: The skill relies on the execution of bash and Python scripts via the shell to perform its core discovery functions, which involves piping arguments and executing binaries based on environmental path resolution.

peer-qa-review