From agentdb-learning
Close the learning loop — record reward signal for an action AgentDB suggested. Use after using anything from agentdb_pattern_search / reflexion_recall / skill_search / learning_route. The bandit needs the signal to improve.
npx claudepluginhub ruvnet/agentdb --plugin agentdb-learningThis skill uses the workspace's default tool permissions.
Close the loop on a memory or routing decision so AgentDB's bandit learns.
Ask the AgentDB bandit which RL algorithm / skill / pattern fits the current task best. Use at task start when there are multiple plausible approaches and you want the data-driven pick.
Retrieve relevant memories for the current task from AgentDB. Use at the start of a task to load prior knowledge, when stuck to surface what worked before, or when the user asks "what do we know about X" / "have we done this before?"
Provides UI/UX resources: 50+ styles, color palettes, font pairings, guidelines, charts for web/mobile across React, Next.js, Vue, Svelte, Tailwind, React Native, Flutter. Aids planning, building, reviewing interfaces.
Share bugs, ideas, or general feedback.
Close the loop on a memory or routing decision so AgentDB's bandit learns.
agentdb_record_feedback(
id: <pattern/skill/episode/decision id>
reward: -1..1
context?: { task, outcome, latency, ... }
)
agentdb_bandit_update(
arm: <bandit arm name>
reward: -1..1
)
| Outcome | Reward |
|---|---|
| Used the suggestion, task succeeded | +1.0 |
| Used the suggestion, task partial success | +0.5 |
| Used the suggestion, didn't help | 0.0 |
| Used the suggestion, made things worse | -0.5 |
| Ignored the suggestion (other reason) | -0.1 (mild downweight) |
| Rejected as wrong / harmful | -1.0 |
const hits = await agentdb_pattern_search(query)
for (const h of useful(hits)) {
use(h)
await agentdb_record_feedback(h.id, +1)
}
for (const h of skipped(hits)) {
await agentdb_record_feedback(h.id, -0.1)
}
Without negative feedback, the bandit only sees "winners" and exploration starves.