oussamadouhou/oh-my-opencode-free-fork
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

fix(hooks): use last assistant message tokens instead of cumulative sum

Browse Source 
Previously, token calculation accumulated ALL assistant messages' tokens,
causing incorrect usage display (e.g., 524.9%) after compaction.
Now uses only the last message's input tokens, which reflects the actual
current context window usage.

🤖 GENERATED WITH ASSISTANCE OF [OhMyOpenCode](https://github.com/code-yeongyu/oh-my-opencode)
This commit is contained in:
YeonGyu-Kim
2025-12-13 01:02:21 +09:00
parent 6a565ee126
commit 3d273ff853
2 changed files with 9 additions and 10 deletions
Download Patch File Download Diff File Expand all files Collapse all files

9
src/hooks/context-window-monitor.ts
View File

@@ -52,11 +52,10 @@ export function createContextWindowMonitorHook(ctx: PluginInput) {
const lastAssistant = assistantMessages[assistantMessages.length - 1] const lastAssistant = assistantMessages[assistantMessages.length - 1]
if (lastAssistant.providerID !== "anthropic") return if (lastAssistant.providerID !== "anthropic") return
const totalInputTokens = assistantMessages.reduce((sum, m) => { // Use only the last assistant message's input tokens
const inputTokens = m.tokens?.input ?? 0 // This reflects the ACTUAL current context window usage (post-compaction)
const cacheReadTokens = m.tokens?.cache?.read ?? 0 const lastTokens = lastAssistant.tokens
return sum + inputTokens + cacheReadTokens const totalInputTokens = (lastTokens?.input ?? 0) + (lastTokens?.cache?.read ?? 0)
}, 0)
const actualUsagePercentage = totalInputTokens / ANTHROPIC_ACTUAL_LIMIT const actualUsagePercentage = totalInputTokens / ANTHROPIC_ACTUAL_LIMIT

10
src/hooks/grep-output-truncator.ts
View File

@@ -98,11 +98,11 @@ export function createGrepOutputTruncatorHook(ctx: PluginInput) {
if (assistantMessages.length === 0) return if (assistantMessages.length === 0) return
const totalInputTokens = assistantMessages.reduce((sum, m) => { // Use only the last assistant message's input tokens
const inputTokens = m.tokens?.input ?? 0 // This reflects the ACTUAL current context window usage (post-compaction)
const cacheReadTokens = m.tokens?.cache?.read ?? 0 const lastAssistant = assistantMessages[assistantMessages.length - 1]
return sum + inputTokens + cacheReadTokens const lastTokens = lastAssistant.tokens
}, 0) const totalInputTokens = (lastTokens?.input ?? 0) + (lastTokens?.cache?.read ?? 0)
const remainingTokens = ANTHROPIC_ACTUAL_LIMIT - totalInputTokens const remainingTokens = ANTHROPIC_ACTUAL_LIMIT - totalInputTokens