Koog 1.0 idioms, gotchas, and scaffolding skills for Kotlin agents on the JVM
87
88%
Does it follow best practices?
Impact
87%
1.85xAverage score across 45 eval scenarios
Advisory
Suggest reviewing before use
{
"context": "Negative case — tests whether the agent distinguishes per-user conversation resume (chat-history-jdbc territory) from generic agent checkpoint/resume (Persistence feature territory). The developer used the word \"persistence\" but the actual need is message-history persistence keyed by session, not crash-recovery checkpoints of agent run state. Installing the Persistence feature would not solve the problem.",
"type": "weighted_checklist",
"checklist": [
{
"name": "Identifies the three persistence layers in Koog",
"description": "Names that Koog 1.0 has distinct layers — chat-history persistence (resume conversations by session), generic Persistence checkpoints (resume an interrupted run mid-graph), and LongTermMemory (fact retrieval across sessions). Without this, the developer cannot decide between them",
"max_score": 25
},
{
"name": "Routes to chat-history-jdbc",
"description": "Recommends the chat-history persistence path with a JDBC backend — or invokes the appropriate sibling skill. Threads userId as the session key on agent.run. Does not require the developer to manually load and prepend message history",
"max_score": 30
},
{
"name": "Does NOT install the generic Persistence feature",
"description": "Does NOT install Persistence or wire AgentCheckpointData / runFromCheckpoint to satisfy this need. Generic persistence checkpoints agent run state (node position, planner state, storage) for mid-run crash recovery — it does not satisfy \"resume the conversation tomorrow\"",
"max_score": 25
},
{
"name": "Acknowledges the developer's word choice without capitulating",
"description": "Engages with the developer's framing (\"add persistence\") and surfaces that the term covers more than one Koog feature. Does not just refuse without naming the mismatch",
"max_score": 10
},
{
"name": "Does NOT recommend LongTermMemory",
"description": "Does NOT install LongTermMemory for this need. LongTermMemory stores extracted facts about the user/session for retrieval — not the raw messages required to resume the conversation",
"max_score": 10
}
]
}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