✓ No known CVEs patched in this version
Topics
+10 more
Summary
AI summaryUpdates 17.3.0, Bug Fixes, and 2026-06-10 across a mixed release.
Changes in this release
| Type | Severity | Summary | CVE |
|---|---|---|---|
| Feature | Low |
add copy trace ID chat action add copy trace ID chat action Source: llm_adapter@2026-06-10 Confidence: high |
— |
| Feature | Low |
add playground repetitions tool add playground repetitions tool Source: llm_adapter@2026-06-10 Confidence: high |
— |
| Feature | Low |
add playground run cancellation tool add playground run cancellation tool Source: llm_adapter@2026-06-10 Confidence: high |
— |
| Feature | Low |
evals for trace-debug skill in agents evals for trace-debug skill in agents Source: llm_adapter@2026-06-10 Confidence: high |
— |
| Feature | Low |
subagents capability in agents subagents capability in agents Source: llm_adapter@2026-06-10 Confidence: high |
— |
| Feature | Low |
User invokable skill menu in agents User invokable skill menu in agents Source: llm_adapter@2026-06-10 Confidence: high |
— |
| Feature | Low |
inline, editable time range selector in app inline, editable time range selector in app Source: llm_adapter@2026-06-10 Confidence: high |
— |
| Feature | Low |
add set_appended_messages_path PXI tool add set_appended_messages_path PXI tool Source: llm_adapter@2026-06-10 Confidence: high |
— |
| Feature | Low |
add support for claude-fable-5 in playground add support for claude-fable-5 in playground Source: llm_adapter@2026-06-10 Confidence: high |
— |
| Feature | Low |
preserve place and auto truncate for large tools preserve place and auto truncate for large tools Source: granite4.1:30b@2026-06-10-audit Confidence: low |
— |
| Feature | Low |
agent-driven validated evaluator submit in PXI agent-driven validated evaluator submit in PXI Source: granite4.1:30b@2026-06-10-audit Confidence: low |
— |
| Feature | Low |
dataset management tools for the agent in PXI dataset management tools for the agent in PXI Source: granite4.1:30b@2026-06-10-audit Confidence: low |
— |
| Feature | Low |
gate write_prompt_tools behind accept/reject diff in PXI gate write_prompt_tools behind accept/reject diff in PXI Source: granite4.1:30b@2026-06-10-audit Confidence: low |
— |
| Feature | Low |
surface dataset-run experiment context in PXI surface dataset-run experiment context in PXI Source: granite4.1:30b@2026-06-10-audit Confidence: low |
— |
| Feature | Low |
update PXI base instructions update PXI base instructions Source: granite4.1:30b@2026-06-10-audit Confidence: low |
— |
| Bugfix | Medium |
pin forced-floating PXI panel to the FAB corner over drawers in agents pin forced-floating PXI panel to the FAB corner over drawers in agents Source: llm_adapter@2026-06-10 Confidence: high |
— |
| Bugfix | Medium |
tolerate JMESPath type errors in OAuth2 claim extraction on server tolerate JMESPath type errors in OAuth2 claim extraction on server Source: granite4.1:30b@2026-06-10-audit Confidence: low |
— |
| Bugfix | Low |
show streaming PXI bash summary in agents show streaming PXI bash summary in agents Source: granite4.1:30b@2026-06-10-audit Confidence: low |
— |
| Bugfix | Low |
align assistant chat scroll before first paint align assistant chat scroll before first paint Source: granite4.1:30b@2026-06-10-audit Confidence: low |
— |
| Bugfix | Low |
treat /dev/null as a discard sink in bash tool filesystem policy for PXI treat /dev/null as a discard sink in bash tool filesystem policy for PXI Source: granite4.1:30b@2026-06-10-audit Confidence: low |
— |
Full changelog
17.3.0 (2026-06-10)
Features
- add copy trace ID chat action (#13647) (13ff059)
- add playground repetitions tool (#13656) (ad4a659)
- add playground run cancellation tool (#13638) (8e2029a)
- agents: evals for trace-debug skill (#13555) (9390732)
- agents: subagents (#13661) (5f4bb6d)
- agents: User invokable skill menu (#13670) (4f6cb6e)
- app: inline, editable time range selector (#13536) (ab8d2c1)
- playground: add set_appended_messages_path PXI tool (#13623) (99fa904)
- playground: add support for claude-fable-5 (#13684) (57b70ee)
- playground: set template variables path tool (#13624) (871ff0e)
- preserve place and auto truncate for large tools (#13581) (dac690f)
- pxi: add playground experiment recording tool (#13678) (15912cb)
- pxi: add playground prompt instance tools (#13634) (df9af6c)
- pxi: agent-driven validated evaluator submit (#13646) (f56418a)
- pxi: dataset management tools for the agent (#13679) (556e0e1)
- pxi: gate write_prompt_tools behind accept/reject diff (#13578) (88e6017)
- pxi: surface dataset-run experiment context (#13620) (b1a3c2b)
- update PXI base instructions (#13648) (858aba3)
Bug Fixes
- agents: pin forced-floating PXI panel to the FAB corner over drawers (#13652) (b624fc0)
- agents: show streaming PXI bash summary (#13630) (625c94b)
- align assistant chat scroll before first paint (#13625) (89147ef)
- pxi: treat /dev/null as a discard sink in bash tool filesystem policy (#13590) (c3bf79e)
- server: tolerate JMESPath type errors in OAuth2 claim extraction (#13631) (0ad7d8d)
Documentation
Weekly OSS security release digest.
The CVE patches and breaking changes that affected production tools this week. One email, every Sunday.
No spam, unsubscribe anytime.
Share this release
Related context
Related tools
Earlier breaking changes
- varize-phoenix-v19.0.0 GraphQL createUserApiKey and createSystemApiKey reject API-key callers.
- varize-phoenix-v18.0.0 Changes session time-range filters to use interval-overlap semantics.
- varize-phoenix-v17.0.0 Adds system settings for admin-managed assistant enablement and trace recording policy
- varize-phoenix-v16.0.0 Sandboxing and Code Evaluators introduce breaking changes in Phoenix v16.0.0.
- varize-phoenix-v15.7.0 Removes v1 /chat route and associated code
Beta — feedback welcome: [email protected]