Reflection
by AaaS · open-source · Last verified 2026-03-01
Allows agents to evaluate their own outputs, identify errors or weaknesses, and iteratively improve responses. Implements self-critique loops where the agent reviews its work against quality criteria and refines until standards are met.
https://aaas.blog/skill/reflection ↗C
C—Below Average
Adoption: BQuality: AFreshness: ACitations: FEngagement: F
Specifications
- License
- MIT
- Pricing
- open-source
- Capabilities
- self-critique, error-identification, iterative-refinement, quality-assessment, confidence-calibration
- Integrations
- langchain, llama-index
- Use Cases
- output-quality-improvement, error-correction, writing-refinement, code-quality-enforcement
- API Available
- No
- Difficulty
- advanced
- Prerequisites
- chain-of-thought
- Supported Agents
- claude-code, devin
- Tags
- reflection, self-evaluation, metacognition, agents, improvement
- Added
- 2026-03-17
- Completeness
- 87%
Index Score
43Adoption
64
Quality
84
Freshness
88
Citations
3
Engagement
0