Fix stale reasoning token counts in budget estimation
Cluster 73624 · agent. Data as of
Nothing open on GitHub. Fix merged in #82116 on 9 Aug 2026. Check it covers your case.
The short answer
- Known?
- Yes. 4 related items grouped here, 1 issue already closed.
- Fix event
- A fix was merged to main in PR #82116. 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.
These members address a bug where _estimate_msg_budget_tokens incorrectly charges stale reasoning blocks against the tail budget, consuming 19–24% of the budget on blocks that are never replayed. The fix ensures actual, current reasoning token counts are used instead of stale values.
Open issues0
Open PRs0
Fixes on record
-
#82116 fix(compression): charge stale thinking to the tail budget only on the newest assistant turn (#73624)
files:
agent/context_compressor.py,tests/agent/test_replay_budget_accounting.py
2 PRs closed without merging
From the close comments: partial fix 1 · superseded by another PR 1
Closed issues · 1
- Closed#73624`_estimate_msg_budget_tokens` charges stale reasoning to the compression tail budget — 19-24% of the budget spent on blocks no adapter replays
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structural links throughout): every member is connected by structural evidence: shared changed files, or a closing/superseded reference.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
0.50: 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
- #82116
- Closed issues
- 1
- Attempts
- 2 PRs closed without merging
- Fix link checked
- 9 Aug 2026