reduction_percent
Method
This is a method of the OptimizationResult class in the models module.Percentage of tokens reduced.
Signature
Returns
The result of the operation.
Source
View on GitHub
praisonaiagents/context/models.py at line 223Guide
Reference
- achat
- aexecute
- Agent ID
- Analyze Prompt
- arun
- astart
- Auto Memory
- background
- chat
- Chat With Context
- checkpoints
- Clean Json Output
- Clear History
- console
- Context Manager
- Delete History
- Delete History Matching
- Display Name
- ephemeral
- execute
- Execute Tool
- Execute Tool Async
- From Template
- Generate Task
- Get Available Tools
- Get History Size
- Get Learn Context
- Get Memory Context
- Get Recommended Stage
- Get Rules Context
- Get Skills Prompt
- Handoff To
- Handoff To Async
- Iter Stream
- launch
- LLM Model
- Output Style
- policy
- Prune History
- query
- RAG
- RAG Query
- Retrieval Config
- retrieve
- Rules Manager
- run
- Run Autonomous
- Run Autonomous Async
- Run Until
- Run Until Async
- Session ID
- Skill Manager
- start
- Store Memory
- Stream Emitter
- Switch Model
- Thinking Budget
- Get App
- serve
- Add Task
- Aexecute Task
- Append To State
- Arun All Tasks
- Arun Task
- astart
- Clean Json Output
- Clear State
- Context Manager
- Current Plan
- Default Completion Checker
- Delete State
- Display Token Usage
- Execute Task
- Get Agent Details
- Get All State
- Get All Tasks Status
- Get Detailed Token Report
- Get Plan Markdown
- Get State
- Get Task Details
- Get Task Result
- Get Task Status
- Get Todo Markdown
- Get Token Usage Summary
- Has State
- Increment State
- launch
- Restore Session State
- run
- Run All Tasks
- Run Task
- Save Output To File
- Save Session State
- Set State
- start
- Todo List
- Update Plan Step Status
- Update State
- execute
- alisten
- asay
- aspeech
- atranscribe
- console
- listen
- litellm
- say
- speech
- transcribe
- Is Available
- Mark Error
- Mark Rate Limited
- reset
- astart
- start
- achat
- chat
- name
- RAG
- run
- Channel ID
- Channel Type
- Is Channel Allowed
- Is User Allowed
- Is Webhook Mode
- command
- Command Args
- Is Command
- text
- channel
- content
- Message ID
- sender
- Bot User
- Delete Message
- Edit Message
- Get Agent
- Get Channel
- Get User
- Is Running
- On Command
- On Message
- platform
- Send Message
- Send Typing
- Set Agent
- start
- stop
- Is Bot
- User ID
- username
- Fixed Total
- Get Segment Budget
- History Budget
- usable
- Supported Chunkers
- chunk
- chunker
- Embedding Model
- aexecute
- agenerate
- console
- execute
- Execute Code
- explain
- fix
- generate
- Generate Code
- litellm
- refactor
- review
- evaluate
- Aanalyze Codebase
- Acreate Implementation Blueprint
- Agenerate Prp
- Analyze Codebase
- Analyze Codebase With Gitingest
- Analyze Integration Points
- Analyze Test Patterns
- Build Implementation Blueprint
- Compile Context Documentation
- Create Implementation Blueprint
- Create Quality Gates
- Create Validation Framework
- Execute Prp
- Extract Implementation Patterns
- Generate Comprehensive Prp
- Generate Feature Prp
- Generate Prp
- Get Agent Interaction Summary
- Log Debug
- Perform Ast Analysis
- Save Comprehensive Session Report
- Save Markdown Output
- Setup Logging
- Setup Output Directories
- start
- For Recipe
- From Env
- merge
- Add Segment
- Get Segment
- Set Segment
- total
- Capture LLM Boundary
- Emergency Truncate
- Estimate Tokens
- Get History
- Get Last Snapshot Hook
- Get Resolved Config
- Get Stats
- Get Tool Budget
- process
- Register Snapshot Callback
- reset
- Set Tool Budget
- Truncate Tool Output
- optimize
- Format For Prompt
- Has Citations
- From Config
- aresearch
- Async Openai Client
- clarify
- Follow Up
- Gemini Client
- Openai Client
- research
- Rewrite Query
- Get All Sources
- Get Citation Text
- evaluate
- Get Target
- aembed
- Aembed Batch
- asimilarity
- console
- embed
- Embed Batch
- Find Most Similar
- litellm
- similarity
- evaluate
- Add Profile
- Get Next Profile
- Get Profile
- Get Retry Delay
- List Profiles
- Mark Failure
- Mark Success
- On Failover
- Remove Profile
- Reset All
- status
- Get Next Profile
- Mark Failure
- Mark Success
- display
- start
- stop
- Client ID
- close
- Connected At
- Is Connected
- receive
- send
- HTTP URL
- Is Secure
- Ws URL
- broadcast
- Close Session
- Create Session
- emit
- Get Agent
- Get Session
- health
- host
- Is Running
- List Agents
- List Sessions
- On Event
- port
- Register Agent
- start
- stop
- Unregister Agent
- Add Message
- Agent ID
- Client ID
- close
- Created At
- Get Messages
- Get State
- Is Active
- Last Activity
- Session ID
- Set State
- Default Tool Description
- Default Tool Name
- Execute Async
- Execute Programmatic
- To Tool Function
- Tool Description
- Tool Name
- matches
- Get Reason
- Is Blocking
- Should Stop
- clear
- Disable Hook
- Enable Hook
- enabled
- Get Hooks
- Has Hooks
- List Hooks
- on
- register
- Register Command
- Register Function
- Set Global Timeout
- unregister
- allow
- ask
- block
- deny
- Is Allowed
- Is Denied
- Aggregate Context
- execute
- Execute Sync
- Get Blocking Reason
- Is Blocked
- registry
- achat
- aedit
- agenerate
- Agenerate Image
- avariation
- chat
- edit
- generate
- Generate Image
- litellm
- variation
- add
- chunker
- config
- delete
- Delete All
- get
- Get All
- Get Corpus Stats
- history
- index
- markdown
- memory
- Normalize Content
- reset
- search
- store
- update
- Get Tools
- shutdown
- To Openai Tool
- Call Tool
- run
- shutdown
- Build Context For Task
- Calculate Quality Metrics
- Compute Quality Score
- Delete Long Term
- Delete Memories
- Delete Memories Matching
- Delete Memory
- Delete Short Term
- Finalize Task Output
- Get All Memories
- Get Learn Context
- learn
- Reset All
- Reset Entity Only
- Reset Long Term
- Reset Short Term
- Reset User Memory
- search
- Search Entity
- Search Long Term
- Search Short Term
- Search User Memory
- Search With Quality
- Store Entity
- Store Long Term
- Store Quality
- Store Short Term
- Store User Memory
- execute
- Execute Model Call
- Execute Tool Call
- Has Model Hooks
- Has Tool Hooks
- Run After Model
- Run After Tool
- Run Before Model
- Run Before Tool
- add
- delete
- Delete All
- get
- Get All
- reset
- search
- update
- Get Agent Manager
- Get Agent Policy
- Get Combined Stats
- Get Session Cache
- Prepare Handoff
- Set Agent Policy
- aextract
- aread
- console
- extract
- litellm
- read
- close
- flush
- On LLM Call
- On Span End
- On Span Start
- On Tool Call
- On Trace End
- On Trace Start
- Reduction Percent
- Add Tool
- Clear Session Permissions
- Is Allowed
- Is Empty
- List Tools
- load
- Remove Tool
- save
- agent
- expand
- Expand Basic
- Expand Creative
- Expand Detailed
- Expand Structured
- Add Abbreviation
- Add Abbreviations
- agent
- rewrite
- Rewrite Basic
- Rewrite Contextual
- Rewrite Hyde
- Rewrite Multi Query
- Rewrite Step Back
- Rewrite Sub Queries
- Format Answer With Citations
- Has Citations
- aconnect
- adisconnect
- Asend Audio
- Asend Text
- connect
- console
- disconnect
- On Audio
- On Error
- On Message
- Receive Loop
- Send Audio
- Send Text
- generous
- minimal
- standard
- Get Strategy
- Get Token Budget
- Should Retrieve
- To Knowledge Config
- To RAG Config
- All Queries
- Primary Query
- Get Target
- docker
- e2b
- subprocess
- cleanup
- execute
- Execute File
- Get Status
- Is Available
- List Files
- Read File
- reset
- Run Command
- Sandbox Type
- start
- stop
- Write File
- output
- success
- permissive
- standard
- strict
- Agent
- Add Knowledge
- Add Memory
- chat
- Clear Memory
- Create Agent
- Get Context
- Get State
- Increment State
- knowledge
- memory
- Restore State
- Save State
- Search Knowledge
- Search Memory
- Send Message
- Set State
- Check And Add
- clear
- Get Stats
- Depends On
- Evaluate When
- Execute Callback
- Execute Callback Sync
- Get Next Task
- Initialize Memory
- Store In Memory
- estimate
- Estimate Messages
- run
- acontent
- adownload
- agenerate
- alist
- aremix
- arun
- astart
- astatus
- Await Completion
- console
- content
- download
- generate
- list
- remix
- run
- start
- status
- Video Module
- Wait For Completion
- aanalyze
- acompare
- adescribe
- Aextract Text
- analyze
- compare
- console
- describe
- Extract Text
- litellm
- Add Approval Requirement
- Add Hook
- Adisplay Error
- Adisplay Generating
- Adisplay Instruction
- Adisplay Interaction
- Adisplay Self Reflection
- Adisplay Tool Call
- aembedding
- After Model
- After Tool
- Apply Config Defaults
- Before Model
- Before Tool
- Categorize Hooks
- Clean Triple Backticks
- Cleanup Telemetry Resources
- Clear Approval Context
- Clear Config Cache
- Configure Default Approvals
- Console Approval Callback
- Create Context Agent
- Create Context Manager
- Create Retrieval Config
- Deduplicate Topics
- Detect URL Scheme
- disable
- Disable Performance Mode
- Disable Telemetry
- Display Error
- Display Generating
- Display Instruction
- Display Interaction
- Display Reasoning Steps
- Display Self Reflection
- Display Tool Call
- Display Working Status
- embedding
- enable
- Enable Performance Mode
- Enable Telemetry
- Encode File To Base64
- Ensure Dir
- Evaluate Condition
- Execute Callback
- Execute Sync Callback
- Force Shutdown Telemetry
- Format Percent
- Get All Paths
- Get Approval Callback
- Get Cache Dir
- Get Checkpoints Dir
- Get Config
- Get Config Path
- Get Data Dir
- Get Default
- Get Default Registry
- Get Defaults Config
- Get Dimensions
- Get Docs Dir
- Get Enabled Plugins
- Get Hook Type
- Get Learn Dir
- Get MCP Auth Path
- Get MCP Dir
- Get Memory Dir
- Get Multimodal Message
- Get Permission Allowlist
- Get Permissions Dir
- Get Plugins Config
- Get Plugins Dir
- Get Project Data Dir
- Get Prp Dir
- Get Risk Level
- Get Rules Dir
- Get Runs Dir
- Get Sessions Dir
- Get Skills Dir
- Get Snapshots Dir
- Get Storage Dir
- Get Summaries Dir
- Get Telemetry
- Get Workflows Dir
- handoff
- Keep Last N Messages
- Remove All Tools
- Remove System Messages
- Has Hook
- Is Already Approved
- Is Approval Required
- Is Enabled
- Is Env Auto Approve
- Is Middleware
- Is Numeric String
- Is Path Like
- Is Plugins Enabled
- Is Policy String
- Is Yaml Approved
- List Plugins
- Make Array Error
- Make Preset Error
- Mark Approved
- Merge Config With Overrides
- Parse Policy String
- Parse URL To Config
- Process Task Context
- Process Video
- Prompt With Handoff Instructions
- Register Approval Callback
- Register Display Callback
- Remove Approval Requirement
- Remove Hook
- Request Approval
- Require Approval
- Reset Yaml Approved Tools
- resolve
- Resolve Autonomy
- Resolve Batch
- Resolve Caching
- Resolve Context
- Resolve Execution
- Resolve Guardrail Policies
- Resolve Guardrails
- Resolve Hooks
- Resolve Knowledge
- Resolve Memory
- Resolve Output
- Resolve Planning
- Resolve Reflection
- Resolve Routing
- Resolve Skills
- Resolve Web
- Set Approval Callback
- Set Default Registry
- Set Permission Allowlist
- Set Yaml Approved Tools
- Suggest Similar
- trace
- Track Workflow
- Validate Config
- Wrap Model Call
- Wrap Tool Call
Functions
Reduction Percent • AI Agent SDK
reduction_percent: Percentage of tokens reduced.
⌘I