Menü
Extracts learnings from completed work, feedback, or failures. Updates a persistent learnings file with capped entries. Use after receiving feedback, fixing bugs, or completing complex tasks.
Mit Codex oder Claude installieren Kopieren Sie diesen Prompt, fügen Sie ihn in Codex, Claude oder einen anderen Assistant ein und lassen Sie die Skill-Seite prüfen und installieren.
Basierend auf der SOC-Berufsklassifikation
| name | vibe-reflect-and-compound |
| description | Extracts learnings from completed work, feedback, or failures. Updates a persistent learnings file with capped entries. Use after receiving feedback, fixing bugs, or completing complex tasks. |
| user-invocable | true |
Every problem solved should produce reusable knowledge. Don't just fix — learn.
Gather input — What happened?
Extract patterns — Ask:
Formulate learnings — Each learning should be:
Update persistent storage — Write to:
docs/learnings.md or LEARNINGS.md)Cap management — Keep max 30 active learnings. When exceeding:
Trigger: [What prompted this reflection]
Learnings Extracted:
[Pattern name]: [Actionable learning]
[Pattern name]: [Actionable learning]
Updated: [Where learnings were saved] Active Count: X/30
Dispatches research agents to find real projects, papers, and documented failures before proposing any feature, architecture, or technology choice. Triggers on creative/design tasks, not on implementation of already-designed specs.
Validates completed work against defined acceptance criteria. Use after completing a task that has specific success criteria defined in issues, specs, or task descriptions.
Generates edge case, failure mode, and spec-driven test cases. Covers boundary values, nil inputs, concurrency, resource exhaustion, malformed data, and requirement-linked traceability tests.
Detects and prevents shortcut rationalization — when the agent is about to skip a step, simplify a requirement, or optimize away a constraint. Triggers proactively when rationalization patterns are noticed.
Provides a persistent cross-session task queue for work that should be done but isn't blocking current tasks. Enables background and future-session work tracking.
Safely integrates commits from parallel agent branches using sequential cherry-pick. Use after parallel work completes in isolated branches or worktrees.