---
slug: "Reflexion"
name: "Reflexion"
type: "Reasoning"
layer: "Mind"
ref: "Reflexion#81eb"
sema_id: "sema:Reflexion#mh:SHA-256:81eb9140703ea95f53556741df731af84c1cada66cc1e0f3a1d88e4de5fe54d8"
tags: ["Mind","Reasoning"]
updated: 2026-11-11
mindscape: "bootstrap"
index_url: https://mindscape.world/scapes/bootstrap.json
url: https://mindscape.world/bootstrap/Reflexion/
related: ["Task","Critique","Goal","Outcome","Plan","Scratchpad","ChainOfThought","DogfoodFirst","MetaCheck","OptimisticSolver","PolymorphicSolver","sema:EvaluatorOptimizer#mh:SHA-256:6c2dd39e3dfae94dee24fb49b8769219134e2ca9bb79e047165d9424614eb729"]
---

# Reflexion

Self-critique and retry after failure

Layer: Mind · Category: Reasoning

## Mechanism

A self-improvement loop where after completing a [Task](https://mindscape.world/scapes/bootstrap/Task.md) attempt, the agent generates linguistic self-critique analyzing what went wrong with the [Plan](https://mindscape.world/scapes/bootstrap/Plan.md) regarding the [Goal](https://mindscape.world/scapes/bootstrap/Goal.md), then uses this reflection as context for a retry. Unlike fine-tuning, the model is frozen—improvement comes from explicit textual feedback stored in an episodic memory buffer. It writes self-critique to a [Scratchpad](https://mindscape.world/scapes/bootstrap/Scratchpad.md) memory buffer, using this feedback to improve the next attempt.

## Invariants

- Memory of error persists
- Self-correction based on past failure

## Preconditions

- Failed trial
- [Outcome](https://mindscape.world/scapes/bootstrap/Outcome.md) signal

## Postconditions

- Improved policy/next trial

## Failure modes

- Shallow Reflection: [Critique](https://mindscape.world/scapes/bootstrap/Critique.md) is generic ('try harder') without actionable insight.
- Reflection Drift: Each retry diverges further from the goal.
- Premature Success: Declaring success before the task is actually solved.

## Builds on

- [Task](https://mindscape.world/scapes/bootstrap/Task.md) (Infrastructure; Accepts)
- [Critique](https://mindscape.world/scapes/bootstrap/Critique.md) (Mind; References)
- [Goal](https://mindscape.world/scapes/bootstrap/Goal.md) (Infrastructure; References)
- [Outcome](https://mindscape.world/scapes/bootstrap/Outcome.md) (Infrastructure; References)
- [Plan](https://mindscape.world/scapes/bootstrap/Plan.md) (Infrastructure; References)
- [Scratchpad](https://mindscape.world/scapes/bootstrap/Scratchpad.md) (Mind; References)

## Used by

- [ChainOfThought](https://mindscape.world/scapes/bootstrap/ChainOfThought.md) (Mind; Composes with)
- [DogfoodFirst](https://mindscape.world/scapes/bootstrap/DogfoodFirst.md) (Mind; References)
- [MetaCheck](https://mindscape.world/scapes/bootstrap/MetaCheck.md) (Mind; References)
- [OptimisticSolver](https://mindscape.world/scapes/bootstrap/OptimisticSolver.md) (Society; Composes with)
- [PolymorphicSolver](https://mindscape.world/scapes/bootstrap/PolymorphicSolver.md) (Mind; Composes with)

## Identity

- Sema ID: `sema:Reflexion#mh:SHA-256:81eb9140703ea95f53556741df731af84c1cada66cc1e0f3a1d88e4de5fe54d8`
- Ref: `Reflexion#81eb`
- Supersedes:
  - `sema:Reflexion#mh:SHA-256:51b91d2cdfeb6aed5ea3293154279d16af2c98b0c473e731066198b5d4d43e7f`
  - `sema:Reflexion#mh:SHA-256:4a467a44a1172f5d7d2119e3a4c3646e36a5f8dea4cdca2fa85bea6f92619d83`
  - `sema:Reflexion#mh:SHA-256:b2b59e0b162b7c5f91b36d991ffb143331568be2d1dbc56218ccfccafd40dfc3`

## Links

- [Pattern page](https://mindscape.world/bootstrap/Reflexion/)
- [View in the interactive graph](https://mindscape.world/bootstrap/#Reflexion)
- [JSON](https://mindscape.world/scapes/bootstrap/Reflexion.json)
- [Index of all patterns](https://mindscape.world/scapes/bootstrap.json)