You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Copy file name to clipboardExpand all lines: README.md
-14Lines changed: 0 additions & 14 deletions
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -41,8 +41,6 @@ DCP uses multiple tools and strategies to reduce context size:
41
41
42
42
**Purge Errors** — Prunes tool inputs for tools that returned errors after a configurable number of turns (default: 4). Error messages are preserved for context, but the potentially large input content is removed. Runs automatically on every request with zero LLM cost.
43
43
44
-
**On Idle Analysis** — Uses a language model to semantically analyze conversation context during idle periods and identify tool outputs that are no longer relevant. Disabled by default (legacy behavior).
45
-
46
44
Your session history is never modified—DCP replaces pruned content with placeholders before sending requests to your LLM.
47
45
48
46
## Impact on Prompt Caching
@@ -118,18 +116,6 @@ DCP uses its own config file:
118
116
// Additional tools to protect from pruning
119
117
"protectedTools": [],
120
118
},
121
-
// (Legacy) Run an LLM to analyze what tool calls are no longer relevant on idle
122
-
"onIdle": {
123
-
"enabled":false,
124
-
// Additional tools to protect from pruning
125
-
"protectedTools": [],
126
-
// Override model for analysis (format: "provider/model")
127
-
// "model": "anthropic/claude-haiku-4-5",
128
-
// Show toast notifications when model selection fails
0 commit comments