single-agent-lightweight-growthlisted

# Skill: Single-Agent Lightweight Growth ## Purpose Use this workflow during normal single-agent execution so that the agent can keep improving without enabling the full multi-agent scheduling system. This is the default growth mode. ## Trigger Use for all normal tasks unless multi-agent long-task scheduling is explicitly enabled. ## Principle Growth should be continuous but cheap. Most small tasks should not produce files, long reviews, or heavy process. ## Workflow 1. Execute the user's task normally. 2. Verify before claiming completion. 3. At the end, briefly check whether the task produced reusable experience. 4. If no meaningful experience exists, do not write memory. 5. If a small lesson exists, mention it as a local note or growth candidate. 6. If a repeated or high-impact pattern appears, propose Standard or Major retrospective. 7. Write long-term memory only after the memory write checklist passes. 8. If the intended long-term memory target is unavailable, keep the candidate as a local note and report that no real long-term write happened. 9. For non-trivial tasks, user corrections, recovery rounds, or failures, include a short user-facing growth result in the final or phase report. ## Lightweight Growth Triggers Record a growth candidate when: - The user corrects the agent. - Verification fails. - A wrong assumption is discovered. - A useful check prevented an error. - A decision rule becomes clear. - The same problem appears repeatedly. - A user explic