preserved-thinking changedbuild-with-claude/preserved-thinking
Nearest release: v2.1.308, published 2 hours after this site recorded the change. Shown because the two are within 24 hours of each other. Nothing here says the release caused the edit.
Recorded here
Lines+8added
Lines−8removed
From line
53
where the diff opens
First seen
1 Sep 2026
this site's first read of the page
Recorded edits29to this page, all time
The whole hunk
from line 53, old and new numbered
/
from line 53
5353
5454With the `thinking-binding-controls-2026-08-01` [beta header](https://platform.claude.com/docs/en/api/beta-headers), the response lists each dropped block in a top-level `input_transformations` array with `reason: "model_binding_mismatch"`:
5555
56```json
56```text wrap
5757{
5858 "input_transformations": [
5959 {
from line 116
116116That setting requires the `thinking-binding-controls-2026-08-01` value in the `anthropic-beta` header.
117117```
118118
119It usually ends with a sentence naming what changed, for example that the `system` prompt or the `tools` list differs from when the block was created. [Troubleshooting thinking](https://platform.claude.com/docs/en/build-with-claude/thinking-troubleshooting#error-thinking-block-signature) describes what that sentence can name.
119It usually ends with a sentence naming what changed, for example that the `system` prompt or the `tools` list differs from when the block was created. For what that sentence can name, see [A 400 error says a thinking block signature is invalid](https://platform.claude.com/docs/en/build-with-claude/thinking-troubleshooting#error-thinking-block-signature) in Troubleshooting thinking.
120120
121121A tampered or undecryptable signature is a different failure. It always returns a 400 (``Invalid `signature` in `thinking` block`` with no sentence about the conversation), and `prefix_mismatch_behavior` doesn't apply to it.
122122
from line 716
716716
717717Log `input_transformations` on every turn of your own integration. When the API drops a block, the entry looks like the following:
718718
719```json
719```text wrap
720720{
721721 "input_transformations": [
722722 {
from line 746
746746
747747On an older account, you can also watch production traffic without opting into enforcement: send the beta header, leave `block_binding` out, and log `input_transformations`. Sending the header alone doesn't change what the model receives. Every thinking block that comes after the content you edited fails the check and gets its own `thinking_mismatch_allowed` entry, with the same `path` and `reason` fields as a `thinking_dropped` entry. Blocks before the edit still pass. An edit to `system` or `tools` comes before every block, so it fails every thinking block in the request. One entry looks like this:
748748
749```json
749```text wrap
750750{
751751 "input_transformations": [
752752 {
from line 1107
11071107
11081108### Send per-turn reminders as turn-scoped system messages
11091109
1110A common prefix edit is the per-turn nudge: a line such as "request independent reads together" or "you haven't updated the user in a while" that your code appends after each batch of tool results. To keep reminders from piling up, send each nudge as a [mid-conversation system message](https://platform.claude.com/docs/en/build-with-claude/mid-conversation-system-messages) with `clear_at: "next_user_message"`, placed after the `tool_result` user message. `clear_at` requires the beta header `mid-conversation-system-clear-at-2026-08-21`. The following `messages` array is the request after two tool calls and their results. `messages[3]` is the previous request's nudge, left in place, and `messages[6]` is this request's copy:
1110A common prefix edit is the per-turn nudge: a line such as "request independent reads together" or "you haven't updated the user in a while" that your code appends after each batch of tool results. To keep reminders from piling up, send each nudge as a [mid-conversation system message](https://platform.claude.com/docs/en/build-with-claude/preserved-thinking#new-instructions) with `clear_at: "next_user_message"`, placed after the `tool_result` user message. `clear_at` requires the beta header `mid-conversation-system-clear-at-2026-08-21`. The following `messages` array is the request after two tool calls and their results. `messages[3]` is the previous request's nudge, left in place, and `messages[6]` is this request's copy:
11111111
11121112```json
11131113[
from line 1165
11651165
11661166You have two ways to use these blocks:
11671167
1168* **Declare every tool up front.** Put every tool the session might need in `tools` on the first request, with `defer_loading: true` on any that shouldn't be offered to the model yet. Then turn tools on and off with `tool_addition` and `tool_removal` blocks that name them.
1168* **Declare every tool up front.** Put every tool the session might need in `tools` on the first request, with `defer_loading: true` on any that you don't want to offer to the model yet. Then turn tools on and off with `tool_addition` and `tool_removal` blocks that name them.
11691169* **Start with a snapshot and add tools as you go.** Put the tools you know about in `tools` on the first request. When you learn about another tool, define it inside a `tool_addition` block instead of editing `tools`.
11701170
11711171Either way, `tools` never changes, so earlier thinking stays valid and the prompt cache still hits, with the one exception noted later in this section.
from line 1475
14751475
14761476The response carries the new assistant turn as usual, plus one `input_transformations` entry for each dropped block. For the history in the diagram, that's the thinking on assistant turns 3 and 4:
14771477
1478```json
1478```text wrap
14791479{
14801480 "input_transformations": [
14811481 {
from line 1544
15441544 </Accordion>
15451545
15461546 <Accordion title="How do I handle instruction files such as AGENTS.md or CLAUDE.md that change mid-session?">
1547 Load them once at session start and keep the top-level `system` prompt and `tools` fixed. When a file changes, append the new version at that point in `messages` instead of editing the original. Use a [mid-conversation system message](https://platform.claude.com/docs/en/build-with-claude/mid-conversation-system-messages) for instructions that come from you as the operator. For file text you treat as untrusted, which shouldn't carry system-prompt authority, put the content in the next `user` turn instead. See [Add instructions with a mid-conversation system message](https://platform.claude.com/docs/en/build-with-claude/preserved-thinking#new-instructions) and [Limitations](https://platform.claude.com/docs/en/build-with-claude/mid-conversation-system-messages#limitations).
1547 Load them once at session start and keep the top-level `system` prompt and `tools` fixed. When a file changes, append the new version at that point in `messages` instead of editing the original. Use a mid-conversation system message for instructions that come from you as the operator. For file text you treat as untrusted, which shouldn't carry system-prompt authority, put the content in the next `user` turn instead. See [Add instructions with a mid-conversation system message](https://platform.claude.com/docs/en/build-with-claude/preserved-thinking#new-instructions) and [Limitations](https://platform.claude.com/docs/en/build-with-claude/mid-conversation-system-messages#limitations).
15481548 </Accordion>
15491549
15501550 <Accordion title="Can I resume a saved session later, after a restart or the next day?">
No line in this hunk matches that.