Fix quadratic cost when streaming large tool-call arguments
Cluster 92207 · agent. Data as of
Nothing open on GitHub. Fix merged in #101906 on 3 Sep 2026. Check it covers your case.
The short answer
- Known?
- Yes. 5 related items grouped here, 1 issue already closed.
- Fix event
- A fix was merged to main in PR #101906. Check it covers your case. No release data is tracked.
- Tracker
- The maintainers' tracker marks this
unresolved. That is separate from the fix event. - Still open
- Nothing open on GitHub.
A bug was identified where streamed tool-call arguments were being rebuilt on every incoming chunk, causing O(n²) performance for large tool calls. These items address the problem by collecting chunks in a list and assembling them in linear time.
Open issues0
Open PRs0
Fixes on record
-
#101906 perf(agent): buffer streamed tool-call argument fragments and join once at materialisation (salvage #92217)
files:
agent/chat_completion_helpers.py,tests/run_agent/test_streaming.py
3 PRs closed without merging
From the close comments: merged via another change 2 · duplicate 1
Closed issues · 1
- Closed#92207[Bug]: streamed tool-call arguments are rebuilt on every chunk, so a big tool call costs the square of its size
How this cluster was grouped
-
Link evidence (
structural links throughout): every member is connected by structural evidence: shared changed files, or a closing/superseded reference. -
Pairwise link density
0.40: the share of item pairs that carry a link, counting a structural link in full and a relevance-only link at 0.36.1.00means every pair is linked.
Pick up a Lab task
Check whether a fix covers the open issues here, with your own agent. Bring back evidence for maintainers.
Open the LabAt a glance
- Subsystem
- agent
- Merged PRs
- #101906
- Closed issues
- 1
- Attempts
- 3 PRs closed without merging
- Fix link checked
- 3 Sep 2026