Unverified50% confidenceFactExact time
The read_tool_output session reader accepts a result ID, a UTF-8 byte offset, and a paging cap of at most 4096 bytes, and returns the five fields text, nextOffset, totalBytes, hasMore, and complete.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/15/2026
First Seen
Valid until: 1/13/2027
Sources
Related Entities
Related Claims
UnverifiedThe paging interface for reading the retained copy uses UTF-8 encoded byte positions as the unit of location rather than line numbers or character counts, and the reader validates and rejects offsets that fall in the middle of a multi-byte character.73% similarUnverifiedThe Read tool has added an allow_large option, which enables reading a text file in its entirety when it exceeds the normal size limit and sufficient context space is available.71% similarUnverifiedScopeCairn's read symbol command returns an excerpt of about 40 lines (including the docstring) rather than the complete file, to reduce token consumption.68% similarUnverifiedtiktoken's README gives an average of about 4 bytes of text per token.67% similarUnverifiedThe schema of the single Read tool in claude-code consumes 7,487 tokens.64% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1038790API
curl https://kongchang.com/api/v1/knowledge/claims/1038790MCP
get_claim(id=1038790)