jbaruch/koog
Koog 1.0 idioms, gotchas, and scaffolding skills for Kotlin agents on the JVM
86
88%
Does it follow best practices?
Impact
86%
1.86xAverage score across 45 eval scenarios
Advisory
Suggest reviewing before use
task.mdevals/scenario-20/
Cap an Agent's Token Spend with Graceful Degradation
Problem/Feature Description
A developer runs a Koog 1.0 agent that processes user-submitted tasks. They've had incidents where a single complex task consumed tens of thousands of tokens and produced a large bill. They want to cap each run at 50,000 tokens.
They explicitly say: "Don't just abort when the cap is hit — try to continue by summarizing the history first, and only fail if summarization can't reclaim enough budget." The agent runs against OpenAI's GPT-4o.
Output Specification
Walk through what to install and configure. Produce the modified agent construction and Gradle changes as a single response, labeled.
evals
scenario-1
scenario-2
scenario-3
scenario-4
scenario-5
scenario-6
scenario-7
scenario-8
scenario-9
scenario-10
scenario-11
scenario-12
scenario-13
scenario-14
scenario-15
scenario-16
scenario-17
scenario-18
scenario-19
scenario-20
scenario-21
scenario-22
scenario-23
scenario-24
scenario-25
scenario-26
scenario-27
scenario-28
scenario-29
scenario-30
scenario-31
scenario-32
scenario-33
scenario-34
scenario-35
scenario-36
scenario-37
scenario-38
scenario-39
scenario-40
scenario-41
scenario-42
scenario-43
scenario-44
scenario-45
skills
add-observability
add-persistence
add-rag
add-structured-output
add-token-budgeting
add-tool
cache-llm-calls
define-prompt
domain-model-subtask-pipeline
references
enable-prompt-caching
handle-agent-events
manage-state
migrate-from-0-x
model-planner-subtasks
persist-chat-history
query-sql-from-agent
scaffold-agent
snapshot-and-restore
test-koog-agents
trace-agent-internals
use-attachments
use-functional-agent
use-llm-node-variants
use-planner
wire-a2a
wire-acp-server
wire-ktor-server
wire-mcp-server
wire-spring-boot