The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses the Bash tool to execute agent-validate run, an autonomous verification suite, with a specified timeout.
[PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection because it extracts and follows instructions from external log files generated during the validation process.
Ingestion points: The agent reads data from files in the validator_logs/ directory, including console logs, check logs, and review JSON files.
Boundary markers: No explicit boundary markers or instructions to ignore embedded commands are used when interpolating content from the logs into the agent's context.
Capability inventory: The skill possesses the ability to execute arbitrary shell commands via Bash, delegate tasks via the Task tool, and update local JSON files.
Sanitization: The skill does not perform sanitization on the --- Fix Instructions --- or --- Fix Skill --- blocks extracted from log files before acting on them.
[COMMAND_EXECUTION]: The skill dynamically invokes other agent skills based on names parsed directly from the validator's log output (Fix Skill: <name>), which could lead to unauthorized tool execution if the logs are tampered with.

validator-run