pre_llm_call context lost during multimodal turns
Cluster 71998 · agent. Data as of
Nothing open on GitHub. Fix merged in #118911 on 22 Sep 2026. Check it covers your case.
The short answer
- Known?
- Yes. 3 related items grouped here, 1 issue already closed.
- Fix event
- A fix was merged to main in PR #118911. Check it covers your case. No release data is tracked.
- Tracker
- The maintainers' tracker marks this
resolved, deciding PR #118911. That is separate from the fix event. - Still open
- Nothing open on GitHub.
These issues and a fix address a bug where plugin or memory context set in pre_llm_call hooks is dropped when the agent processes multimodal (image) turns. The work covers detection, a fix, and related concerns around persistence, compaction, and resume.
Open issues0
Open PRs0
Fixes on record
-
#118911 pre_llm_call context reaches and survives multimodal turns: wire, persistence, compaction, resume (#71998, base #72026)
files:
agent/turn_context.py,tests/agent/test_api_content_sidecar.py,website/docs/user-guide/features/hooks.md
1 PR closed without merging
From the close comments: merged via another change 1
-
#72026 fix(agent): deliver pre_llm_call/memory context on multimodal turns (#71998) · merged via another change →
bb87e6a,#118911,#71998
Closed issues · 1
- Closed#71998pre_llm_call plugin context is dropped for multimodal image turns
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structural links throughout): every member is connected by structural evidence: shared changed files, or a closing/superseded reference.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
0.67: 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
- #118911
- Closed issues
- 1
- Attempts
- 1 PR closed without merging
- Fix link checked
- 22 Sep 2026