specialist-injection

Specialist injection dynamically recommends and adds team members when recurring feedback themes emerge. Use recommend-specialist.sh with your current team, feedback themes, and a recurring-count threshold to inject specialists matching identified patterns without removing existing members.

Classification extracts task_type, domains, and recommended agents from CFN Loop task descriptions. This determines loop configuration, iteration count, and specialist roles needed—useful for multi-agent coordination, infrastructure scenarios, and standardizing outputs for CLI tooling and audit trails.

Yes. Specialist injection applies non-destructive team modifications, adding new specialists based on feedback themes while preserving your current team structure and roles.

Use specialist injection when you encounter recurring feedback themes, need multi-agent coordination across different problem domains, or want to scale team composition based on actual task complexity and emerging skill gaps rather than pre-defining the entire team upfront.

Install jq for script execution. Prepare your current Loop 3 team roster, identify recurring feedback themes from prior task iterations, and decide the recurring-count threshold that triggers specialist injection into your team.

Recommendations output standardized specialist profiles and team configurations ready for CLI tooling. Use these outputs to update your Loop 3 loop setup, persist recommendations to audit logs, and cache configurations for repeated task patterns.