The Agent Skills Directory

[SAFE]: The skill's primary function is to provide a standardized method for task state persistence and recovery. The implementation relies on local file operations and does not exhibit any malicious behaviors.
[COMMAND_EXECUTION]: The skill requires access to the Bash tool. The provided instructions use this tool for local environment setup (symlinking) and managing coordination logic, which aligns with the skill's stated purpose of managing complex tasks.
[INDIRECT_PROMPT_INJECTION]: The skill includes functionality to read from local state files to recover task context.
Ingestion points: Reads state files from paths like docs/<topic>-state.md (SKILL.md, Step 3).
Boundary markers: The skill suggests a structured Markdown template for state files, providing implicit structural boundaries, though no explicit 'ignore instructions' delimiters are mandated.
Capability inventory: The skill uses Bash, Write, Edit, Read, Grep, and Glob tools.
Sanitization: No specific sanitization of the state file content is described.
Risk Assessment: The risk is considered minimal as the state files are intended to be managed by the agent itself within the local workspace.

long-task-coordinator