fix(agent): restore be_output and test DeepL error return#16363
Merged
yuzhichang merged 7 commits intoJun 28, 2026
Merged
Conversation
dosubot
Bot
added
size:XS
Contributor
Author
|
Could a maintainer add the |
Contributor
|
No actionable comments were generated in the recent review. 🎉 ℹ️ Recent review info⚙️ Run configurationConfiguration used: Organization UI Review profile: CHILL Plan: Pro Run ID: 📒 Files selected for processing (1)
💤 Files with no reviewable changes (1)
📝 WalkthroughWalkthroughComponentBase adds a helper that wraps a value in a one-row pandas DataFrame with a ChangesDeepL output helper and failure test
Estimated code review effort🎯 2 (Simple) | ⏱️ ~10 minutes Possibly related issues
Possibly related PRs
Suggested labels
Suggested reviewers
Poem
🚥 Pre-merge checks | ✅ 4 | ❌ 1❌ Failed checks (1 warning)
✅ Passed checks (4 passed)
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
Contributor
There was a problem hiding this comment.
🧹 Nitpick comments (1)
test/unit_test/agent/component/test_deepl.py (1)
62-62: 📐 Maintainability & Code Quality | 🔵 Trivial | 💤 Low valueRedundant input setup.
get_inputis patched at Line 64 to return{"content": ["hello"]}, so seedingcpn._param.inputshere has no effect on the code path under test. Drop it to avoid implying it's exercised.♻️ Proposed cleanup
def test_run_returns_error_on_translation_failure(): cpn = _deepl() - cpn._param.inputs = {"content": {"value": ["hello"]}} with patch.object(DeepL, "get_input", return_value={"content": ["hello"]}):🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the rest with a brief reason, keep changes minimal, and validate. In `@test/unit_test/agent/component/test_deepl.py` at line 62, Remove the redundant test setup in test_deepl.py by deleting the cpn._param.inputs assignment, since get_input is already patched to return the needed {"content": ["hello"]}. Keep the test focused on the patched get_input path and rely on that mocked value instead of seeding internal state on cpn.
🤖 Prompt for all review comments with AI agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
Nitpick comments:
In `@test/unit_test/agent/component/test_deepl.py`:
- Line 62: Remove the redundant test setup in test_deepl.py by deleting the
cpn._param.inputs assignment, since get_input is already patched to return the
needed {"content": ["hello"]}. Keep the test focused on the patched get_input
path and rely on that mocked value instead of seeding internal state on cpn.
ℹ️ Review info
⚙️ Run configuration
Configuration used: Organization UI
Review profile: CHILL
Plan: Pro
Run ID: a9bc01b6-d235-47d4-a89d-8c19917bb319
📒 Files selected for processing (2)
agent/component/base.pytest/unit_test/agent/component/test_deepl.py
Contributor
Author
|
Fixed the preflight Ruff failure in 2dd8d93 by removing the unused Verification:
|
Codecov Report✅ All modified and coverable lines are covered by tests. Additional details and impacted files@@ Coverage Diff @@
## main #16363 +/- ##
==========================================
- Coverage 94.56% 93.16% -1.40%
==========================================
Files 10 10
Lines 717 717
Branches 118 118
==========================================
- Hits 678 668 -10
- Misses 25 29 +4
- Partials 14 20 +6 ☔ View full report in Codecov by Harness. 🚀 New features to boost your workflow:
|
yuzhichang
reviewed
Jun 28, 2026
yuzhichang
left a comment
yuzhichang
left a comment
Member
There was a problem hiding this comment.
I re-reviewed the current head locally. The be_output() restoration and the DeepL failure-path regression test both look correct, and I do not see any remaining blocking issue in the current diff. The only unresolved bot comment is a very low-value cleanup in the test setup, so from my side this is ready to merge.
yuzhichang
added a commit
that referenced
this pull request
Jun 29, 2026
## Summary #16332 fixed the missing `return` in DeepL's except branch, but `ComponentBase.be_output` was removed during the agent refactor (#9113) while several components still call it. DeepL (and other tools) would raise `AttributeError` before any error message could be returned. - Restore `ComponentBase.be_output` as `pd.DataFrame([{"content": v}])` (same as pre-refactor behavior) - Add regression test that `_run` returns the `**Error**:` message when translation fails Related to #16329 ## Test plan - [x] `test_run_returns_error_on_translation_failure` - [x] Existing `test_deepl.py` check() tests still pass --------- Co-authored-by: Harsh Kashyap <harshkashyap@Harshs-MacBook-Pro.local> Co-authored-by: Zhichang Yu <yuzhichang@gmail.com>
yuzhichang
added a commit
to yuzhichang/ragflow
that referenced
this pull request
Jun 29, 2026
Ports 5 S-effort PRs from .claude/plans/review-and-port-py-agent-changes.md Execution snapshot (all merged upstream) to the Go agent side. Group 1 - tool ports - infiniflow#15609 ExeSQL SSRF guard: add ValidateDBHost to tool/ssrf.go and wire into exesql.InvokableRun. Loopback, link-local, RFC1918, and empty host are rejected before any driver dispatch. Allowed hosts are dialed via the validated/resolved public IP (DNS-pin). - infiniflow#15436 HTTP timeout: Go side already routes github/qweather/tushare through HTTPHelper (client.Timeout: 30s + per-stage timeouts) and jin10 is a stub. Added regression guard tests: default client timeout, stage timeouts, and AST scan for bare http.Get/Post calls outside the helper. Group 2 - component ports - infiniflow#16363 be_output restore + DeepL error: add BeOutput(v) helper returning {"content": v} frame; add DeepL translation-failure regression test using rewriteHostTransport (avoids mutating package globals which raced TestDeepL_BuildRequest). - infiniflow#15644 Switch no longer matches empty condition: evaluateGroup empty-clauses short-circuit now returns false (was true) so an "and" condition with no clauses falls through to the next condition or end_cpn_ids instead of routing to the empty group's "to" target. Mirrors the all([]) is True fix. Group 4 - API ports - infiniflow#15374 session_id bind to path agent_id: Go DAO already enforces WHERE id = ? AND dialog_id = ? (stronger than the Python post-fetch check). Added 3 IDOR regression tests at the service level (Get/Get mismatch/Delete) confirming the binding holds. Total: 32 new tests across 4 new test files. All agent/tool and agent/component tests pass (320 + 331). All new service IDOR tests pass (3/3). The one failing test in agent/component (TestAgent_ReActExhaustsSteps) is a pre-existing flaky test that also fails on the baseline without my changes. Refs: .claude/plans/review-and-port-py-agent-changes.md Refs: infiniflow#15609 infiniflow#15436 infiniflow#16363 infiniflow#15644 infiniflow#15374
yuzhichang
added a commit
to yuzhichang/ragflow
that referenced
this pull request
Jun 29, 2026
Ports 5 S-effort PRs from .claude/plans/review-and-port-py-agent-changes.md Execution snapshot (all merged upstream) to the Go agent side. Group 1 - tool ports - infiniflow#15609 ExeSQL SSRF guard: add ValidateDBHost to tool/ssrf.go and wire into exesql.InvokableRun. Loopback, link-local, RFC1918, and empty host are rejected before any driver dispatch. Allowed hosts are dialed via the validated/resolved public IP (DNS-pin). - infiniflow#15436 HTTP timeout: Go side already routes github/qweather/tushare through HTTPHelper (client.Timeout: 30s + per-stage timeouts) and jin10 is a stub. Added regression guard tests: default client timeout, stage timeouts, and AST scan for bare http.Get/Post calls outside the helper. Group 2 - component ports - infiniflow#16363 be_output restore + DeepL error: add BeOutput(v) helper returning {"content": v} frame; add DeepL translation-failure regression test using rewriteHostTransport (avoids mutating package globals which raced TestDeepL_BuildRequest). - infiniflow#15644 Switch no longer matches empty condition: evaluateGroup empty-clauses short-circuit now returns false (was true) so an "and" condition with no clauses falls through to the next condition or end_cpn_ids instead of routing to the empty group's "to" target. Mirrors the all([]) is True fix. Group 4 - API ports - infiniflow#15374 session_id bind to path agent_id: Go DAO already enforces WHERE id = ? AND dialog_id = ? (stronger than the Python post-fetch check). Added 3 IDOR regression tests at the service level (Get/Get mismatch/Delete) confirming the binding holds. Total: 32 new tests across 4 new test files. All agent/tool and agent/component tests pass (320 + 331). All new service IDOR tests pass (3/3). The one failing test in agent/component (TestAgent_ReActExhaustsSteps) is a pre-existing flaky test that also fails on the baseline without my changes. Refs: .claude/plans/review-and-port-py-agent-changes.md Refs: infiniflow#15609 infiniflow#15436 infiniflow#16363 infiniflow#15644 infiniflow#15374
yuzhichang
added a commit
to yuzhichang/ragflow
that referenced
this pull request
Jun 29, 2026
Ports 5 S-effort PRs from .claude/plans/review-and-port-py-agent-changes.md Execution snapshot (all merged upstream) to the Go agent side. Group 1 - tool ports - infiniflow#15609 ExeSQL SSRF guard: add ValidateDBHost to tool/ssrf.go and wire into exesql.InvokableRun. Loopback, link-local, RFC1918, and empty host are rejected before any driver dispatch. Allowed hosts are dialed via the validated/resolved public IP (DNS-pin). - infiniflow#15436 HTTP timeout: Go side already routes github/qweather/tushare through HTTPHelper (client.Timeout: 30s + per-stage timeouts) and jin10 is a stub. Added regression guard tests: default client timeout, stage timeouts, and AST scan for bare http.Get/Post calls outside the helper. Group 2 - component ports - infiniflow#16363 be_output restore + DeepL error: add BeOutput(v) helper returning {"content": v} frame; add DeepL translation-failure regression test using rewriteHostTransport (avoids mutating package globals which raced TestDeepL_BuildRequest). - infiniflow#15644 Switch no longer matches empty condition: evaluateGroup empty-clauses short-circuit now returns false (was true) so an "and" condition with no clauses falls through to the next condition or end_cpn_ids instead of routing to the empty group's "to" target. Mirrors the all([]) is True fix. Group 4 - API ports - infiniflow#15374 session_id bind to path agent_id: Go DAO already enforces WHERE id = ? AND dialog_id = ? (stronger than the Python post-fetch check). Added 3 IDOR regression tests at the service level (Get/Get mismatch/Delete) confirming the binding holds. Total: 32 new tests across 4 new test files. All agent/tool and agent/component tests pass (320 + 331). All new service IDOR tests pass (3/3). The one failing test in agent/component (TestAgent_ReActExhaustsSteps) is a pre-existing flaky test that also fails on the baseline without my changes. Refs: .claude/plans/review-and-port-py-agent-changes.md Refs: infiniflow#15609 infiniflow#15436 infiniflow#16363 infiniflow#15644 infiniflow#15374
yuzhichang
added a commit
to yuzhichang/ragflow
that referenced
this pull request
Jun 29, 2026
Ports 5 S-effort PRs from .claude/plans/review-and-port-py-agent-changes.md Execution snapshot (all merged upstream) to the Go agent side. Group 1 - tool ports - infiniflow#15609 ExeSQL SSRF guard: add ValidateDBHost to tool/ssrf.go and wire into exesql.InvokableRun. Loopback, link-local, RFC1918, and empty host are rejected before any driver dispatch. Allowed hosts are dialed via the validated/resolved public IP (DNS-pin). - infiniflow#15436 HTTP timeout: Go side already routes github/qweather/tushare through HTTPHelper (client.Timeout: 30s + per-stage timeouts) and jin10 is a stub. Added regression guard tests: default client timeout, stage timeouts, and AST scan for bare http.Get/Post calls outside the helper. Group 2 - component ports - infiniflow#16363 be_output restore + DeepL error: add BeOutput(v) helper returning {"content": v} frame; add DeepL translation-failure regression test using rewriteHostTransport (avoids mutating package globals which raced TestDeepL_BuildRequest). - infiniflow#15644 Switch no longer matches empty condition: evaluateGroup empty-clauses short-circuit now returns false (was true) so an "and" condition with no clauses falls through to the next condition or end_cpn_ids instead of routing to the empty group's "to" target. Mirrors the all([]) is True fix. Group 4 - API ports - infiniflow#15374 session_id bind to path agent_id: Go DAO already enforces WHERE id = ? AND dialog_id = ? (stronger than the Python post-fetch check). Added 3 IDOR regression tests at the service level (Get/Get mismatch/Delete) confirming the binding holds. Total: 32 new tests across 4 new test files. All agent/tool and agent/component tests pass (320 + 331). All new service IDOR tests pass (3/3). The one failing test in agent/component (TestAgent_ReActExhaustsSteps) is a pre-existing flaky test that also fails on the baseline without my changes. Refs: .claude/plans/review-and-port-py-agent-changes.md Refs: infiniflow#15609 infiniflow#15436 infiniflow#16363 infiniflow#15644 infiniflow#15374
yuzhichang
added a commit
to yuzhichang/ragflow
that referenced
this pull request
Jun 29, 2026
Ports 5 S-effort PRs from .claude/plans/review-and-port-py-agent-changes.md Execution snapshot (all merged upstream) to the Go agent side. Group 1 - tool ports - infiniflow#15609 ExeSQL SSRF guard: add ValidateDBHost to tool/ssrf.go and wire into exesql.InvokableRun. Loopback, link-local, RFC1918, and empty host are rejected before any driver dispatch. Allowed hosts are dialed via the validated/resolved public IP (DNS-pin). - infiniflow#15436 HTTP timeout: Go side already routes github/qweather/tushare through HTTPHelper (client.Timeout: 30s + per-stage timeouts) and jin10 is a stub. Added regression guard tests: default client timeout, stage timeouts, and AST scan for bare http.Get/Post calls outside the helper. Group 2 - component ports - infiniflow#16363 be_output restore + DeepL error: add BeOutput(v) helper returning {"content": v} frame; add DeepL translation-failure regression test using rewriteHostTransport (avoids mutating package globals which raced TestDeepL_BuildRequest). - infiniflow#15644 Switch no longer matches empty condition: evaluateGroup empty-clauses short-circuit now returns false (was true) so an "and" condition with no clauses falls through to the next condition or end_cpn_ids instead of routing to the empty group's "to" target. Mirrors the all([]) is True fix. Group 4 - API ports - infiniflow#15374 session_id bind to path agent_id: Go DAO already enforces WHERE id = ? AND dialog_id = ? (stronger than the Python post-fetch check). Added 3 IDOR regression tests at the service level (Get/Get mismatch/Delete) confirming the binding holds. Total: 32 new tests across 4 new test files. All agent/tool and agent/component tests pass (320 + 331). All new service IDOR tests pass (3/3). The one failing test in agent/component (TestAgent_ReActExhaustsSteps) is a pre-existing flaky test that also fails on the baseline without my changes. Refs: .claude/plans/review-and-port-py-agent-changes.md Refs: infiniflow#15609 infiniflow#15436 infiniflow#16363 infiniflow#15644 infiniflow#15374
yuzhichang
added a commit
to yuzhichang/ragflow
that referenced
this pull request
Jun 30, 2026
Ports 5 S-effort PRs from .claude/plans/review-and-port-py-agent-changes.md Execution snapshot (all merged upstream) to the Go agent side. Group 1 - tool ports - infiniflow#15609 ExeSQL SSRF guard: add ValidateDBHost to tool/ssrf.go and wire into exesql.InvokableRun. Loopback, link-local, RFC1918, and empty host are rejected before any driver dispatch. Allowed hosts are dialed via the validated/resolved public IP (DNS-pin). - infiniflow#15436 HTTP timeout: Go side already routes github/qweather/tushare through HTTPHelper (client.Timeout: 30s + per-stage timeouts) and jin10 is a stub. Added regression guard tests: default client timeout, stage timeouts, and AST scan for bare http.Get/Post calls outside the helper. Group 2 - component ports - infiniflow#16363 be_output restore + DeepL error: add BeOutput(v) helper returning {"content": v} frame; add DeepL translation-failure regression test using rewriteHostTransport (avoids mutating package globals which raced TestDeepL_BuildRequest). - infiniflow#15644 Switch no longer matches empty condition: evaluateGroup empty-clauses short-circuit now returns false (was true) so an "and" condition with no clauses falls through to the next condition or end_cpn_ids instead of routing to the empty group's "to" target. Mirrors the all([]) is True fix. Group 4 - API ports - infiniflow#15374 session_id bind to path agent_id: Go DAO already enforces WHERE id = ? AND dialog_id = ? (stronger than the Python post-fetch check). Added 3 IDOR regression tests at the service level (Get/Get mismatch/Delete) confirming the binding holds. Total: 32 new tests across 4 new test files. All agent/tool and agent/component tests pass (320 + 331). All new service IDOR tests pass (3/3). The one failing test in agent/component (TestAgent_ReActExhaustsSteps) is a pre-existing flaky test that also fails on the baseline without my changes. Refs: .claude/plans/review-and-port-py-agent-changes.md Refs: infiniflow#15609 infiniflow#15436 infiniflow#16363 infiniflow#15644 infiniflow#15374 Apply unstaged agent-go-port prep work docker/.env - switch API_PROXY_SCHEME from python to go (Go server is the default) - enable SANDBOX_ENABLED=1 and extend COMPOSE_PROFILES with ragflow-go web/CLAUDE.md - document frontend agent conventions: CSS/flex layout debugging, scope boundaries, i18n language files, query-key factory pattern, request layering (hook -> service -> next-request), shared UI lock, utility libs (dayjs, ahooks, lodash) web/.agents/skills/tanstack-query-best-practices/ - remove the full tanstack-query-best-practices skill set (SKILL.md + 21 rule files, 1,535 lines total) Lefthook pre-commit was satisfied after stripping two pre-existing trailing-whitespace lines from docker/.env comments. chore(hooks): enable auto-fix for staged file checks Convert lefthook pre-commit to B-mode (check + auto-fix in place) for the four fixable jobs: - end-of-file-fixer, trailing-whitespace, mixed-line-ending: these wrap tools/hooks/check_files.py, which gains a --fix mode that rewrites staged files in place. Lefthook re-stages the fixes via stage_fixed: true, so the next commit naturally includes the cleanup. The user no longer has to manually strip whitespace, append newlines, or normalize line endings. - web-eslint: ESLint --fix is enabled; safe auto-fixes are applied, warnings still surface. trailing-whitespace now uses a regex ([ \t]+(?=\r?\n|$)) so it correctly detects/strips whitespace before a line terminator and preserves CRLF when only stripping whitespace. A previous splitlines-based rstrip pass would silently miss whitespace-before-newline. Checks that only validate (json, yaml) or detect states that need human review (case-conflict, merge-conflict, broken symlinks) remain check-only; their --fix flag is accepted but a no-op. ruff check/format, gofmt, and web-prettier already auto-fixed in the previous config and were left unchanged. fix
yuzhichang
added a commit
to yuzhichang/ragflow
that referenced
this pull request
Jun 30, 2026
Ports 5 S-effort PRs from .claude/plans/review-and-port-py-agent-changes.md Execution snapshot (all merged upstream) to the Go agent side. Group 1 - tool ports - infiniflow#15609 ExeSQL SSRF guard: add ValidateDBHost to tool/ssrf.go and wire into exesql.InvokableRun. Loopback, link-local, RFC1918, and empty host are rejected before any driver dispatch. Allowed hosts are dialed via the validated/resolved public IP (DNS-pin). - infiniflow#15436 HTTP timeout: Go side already routes github/qweather/tushare through HTTPHelper (client.Timeout: 30s + per-stage timeouts) and jin10 is a stub. Added regression guard tests: default client timeout, stage timeouts, and AST scan for bare http.Get/Post calls outside the helper. Group 2 - component ports - infiniflow#16363 be_output restore + DeepL error: add BeOutput(v) helper returning {"content": v} frame; add DeepL translation-failure regression test using rewriteHostTransport (avoids mutating package globals which raced TestDeepL_BuildRequest). - infiniflow#15644 Switch no longer matches empty condition: evaluateGroup empty-clauses short-circuit now returns false (was true) so an "and" condition with no clauses falls through to the next condition or end_cpn_ids instead of routing to the empty group's "to" target. Mirrors the all([]) is True fix. Group 4 - API ports - infiniflow#15374 session_id bind to path agent_id: Go DAO already enforces WHERE id = ? AND dialog_id = ? (stronger than the Python post-fetch check). Added 3 IDOR regression tests at the service level (Get/Get mismatch/Delete) confirming the binding holds. Total: 32 new tests across 4 new test files. All agent/tool and agent/component tests pass (320 + 331). All new service IDOR tests pass (3/3). The one failing test in agent/component (TestAgent_ReActExhaustsSteps) is a pre-existing flaky test that also fails on the baseline without my changes. Refs: .claude/plans/review-and-port-py-agent-changes.md Refs: infiniflow#15609 infiniflow#15436 infiniflow#16363 infiniflow#15644 infiniflow#15374 Apply unstaged agent-go-port prep work docker/.env - switch API_PROXY_SCHEME from python to go (Go server is the default) - enable SANDBOX_ENABLED=1 and extend COMPOSE_PROFILES with ragflow-go web/CLAUDE.md - document frontend agent conventions: CSS/flex layout debugging, scope boundaries, i18n language files, query-key factory pattern, request layering (hook -> service -> next-request), shared UI lock, utility libs (dayjs, ahooks, lodash) web/.agents/skills/tanstack-query-best-practices/ - remove the full tanstack-query-best-practices skill set (SKILL.md + 21 rule files, 1,535 lines total) Lefthook pre-commit was satisfied after stripping two pre-existing trailing-whitespace lines from docker/.env comments. chore(hooks): enable auto-fix for staged file checks Convert lefthook pre-commit to B-mode (check + auto-fix in place) for the four fixable jobs: - end-of-file-fixer, trailing-whitespace, mixed-line-ending: these wrap tools/hooks/check_files.py, which gains a --fix mode that rewrites staged files in place. Lefthook re-stages the fixes via stage_fixed: true, so the next commit naturally includes the cleanup. The user no longer has to manually strip whitespace, append newlines, or normalize line endings. - web-eslint: ESLint --fix is enabled; safe auto-fixes are applied, warnings still surface. trailing-whitespace now uses a regex ([ \t]+(?=\r?\n|$)) so it correctly detects/strips whitespace before a line terminator and preserves CRLF when only stripping whitespace. A previous splitlines-based rstrip pass would silently miss whitespace-before-newline. Checks that only validate (json, yaml) or detect states that need human review (case-conflict, merge-conflict, broken symlinks) remain check-only; their --fix flag is accepted but a no-op. ruff check/format, gofmt, and web-prettier already auto-fixed in the previous config and were left unchanged.
yuzhichang
added a commit
that referenced
this pull request
Jun 30, 2026
) Mirrors 14 merged upstream PRs into the Go agent port. PRs ported: - #15609 ExeSQL SSRF guard + DNS pin - #15436 HTTP timeout on external API tools - #16363 be_output restore + DeepL error path - #15644 switch no longer matches empty condition - #15374 session_id bind to path agent_id (DAO idor guard) - #16169 sandbox artifact ownership gate - #15457 tenant ownership on agentbots - #15145 rerun agent document access check - #15446 thinking switch (component portion; provider policy lives in internal/llm) - #15426 Invoke URL/proxy SSRF + DNS pin + no-redirects - #15238 agentbot thinking-logs beta endpoint - #14589 UserFillUp SSE event propagation - #14890 anonymous webhook opt-in - #15068 PipelineChunker new component (text/file_ref/parser_id dispatch; file-format extraction is a follow-up) 40 files, +2355 / -58 lines. 33 new tests, all targeted package suites pass (1721 + 4 skipped); 1 pre-existing flaky test unrelated.
yuzhichang
added a commit
to ondery/ragflow
that referenced
this pull request
Jul 1, 2026
…sion/user + input/output to Langfuse for agent runs Rebuilt on current upstream/main after force-push; cherry-pick of the 7 PR-original commits collapsed into a single net diff to avoid conflicts with intervening commits pulled in by repeated merges of upstream/main. Original PR work (commits dropped during rebuild because their content already lives on upstream/main or was a merge of upstream/main): c7339ae Enhance LLMService with token usage reporting 0266cf0 Implement token usage tracking and Langfuse integration ab2b267 Include session_id in run_kwargs 8bf937c Implement token usage tracking in chat model c955988 Enhance token tracking with context variables and functions 3e88ed5 fix(agent): address CodeRabbit review on token usage tracking 74b8c8c fix(agent): address CodeRabbit follow-up review Forced-upstream duplicates dropped (already merged to main): 70546ea (infiniflow#16396), 5b09910 (infiniflow#16403), 5d7f0fd (infiniflow#16405), 730f33b (infiniflow#16407), a06343e (infiniflow#16408), f90be41 (infiniflow#16410), c4fe68e (infiniflow#16409), d4a0e71 (infiniflow#14773), ef42fd7 (infiniflow#15436), ab1015d (infiniflow#16363), db188cc (infiniflow#15446), 56ece68 (infiniflow#15609), 2589b6e (infiniflow#15644), 74c2944 (infiniflow#15374), 78832ff (infiniflow#14985), bf18b59 (infiniflow#15457), 6c87225 (infiniflow#14890), e16d1a0 (infiniflow#15426), 4712fcb (infiniflow#16169), 0536233 (infiniflow#14589), 08c807a (infiniflow#15145) Plus 18 'Merge branch main into main' commits dropped (polluted topology).
yuzhichang
added a commit
to ondery/ragflow
that referenced
this pull request
Jul 1, 2026
…sion/user + input/output to Langfuse for agent runs Rebuilt on current upstream/main after force-push; cherry-pick of the 7 PR-original commits collapsed into a single net diff to avoid conflicts with intervening commits pulled in by repeated merges of upstream/main. Original PR work (commits dropped during rebuild because their content already lives on upstream/main or was a merge of upstream/main): c7339ae Enhance LLMService with token usage reporting 0266cf0 Implement token usage tracking and Langfuse integration ab2b267 Include session_id in run_kwargs 8bf937c Implement token usage tracking in chat model c955988 Enhance token tracking with context variables and functions 3e88ed5 fix(agent): address CodeRabbit review on token usage tracking 74b8c8c fix(agent): address CodeRabbit follow-up review Forced-upstream duplicates dropped (already merged to main): 70546ea (infiniflow#16396), 5b09910 (infiniflow#16403), 5d7f0fd (infiniflow#16405), 730f33b (infiniflow#16407), a06343e (infiniflow#16408), f90be41 (infiniflow#16410), c4fe68e (infiniflow#16409), d4a0e71 (infiniflow#14773), ef42fd7 (infiniflow#15436), ab1015d (infiniflow#16363), db188cc (infiniflow#15446), 56ece68 (infiniflow#15609), 2589b6e (infiniflow#15644), 74c2944 (infiniflow#15374), 78832ff (infiniflow#14985), bf18b59 (infiniflow#15457), 6c87225 (infiniflow#14890), e16d1a0 (infiniflow#15426), 4712fcb (infiniflow#16169), 0536233 (infiniflow#14589), 08c807a (infiniflow#15145) Plus 18 'Merge branch main into main' commits dropped (polluted topology). Co-authored-by: yzc <yuzhichang@gmail.com>
xugangqiang
pushed a commit
to xugangqiang/ragflow
that referenced
this pull request
Jul 1, 2026
…iniflow#16455) Mirrors 14 merged upstream PRs into the Go agent port. PRs ported: - infiniflow#15609 ExeSQL SSRF guard + DNS pin - infiniflow#15436 HTTP timeout on external API tools - infiniflow#16363 be_output restore + DeepL error path - infiniflow#15644 switch no longer matches empty condition - infiniflow#15374 session_id bind to path agent_id (DAO idor guard) - infiniflow#16169 sandbox artifact ownership gate - infiniflow#15457 tenant ownership on agentbots - infiniflow#15145 rerun agent document access check - infiniflow#15446 thinking switch (component portion; provider policy lives in internal/llm) - infiniflow#15426 Invoke URL/proxy SSRF + DNS pin + no-redirects - infiniflow#15238 agentbot thinking-logs beta endpoint - infiniflow#14589 UserFillUp SSE event propagation - infiniflow#14890 anonymous webhook opt-in - infiniflow#15068 PipelineChunker new component (text/file_ref/parser_id dispatch; file-format extraction is a follow-up) 40 files, +2355 / -58 lines. 33 new tests, all targeted package suites pass (1721 + 4 skipped); 1 pre-existing flaky test unrelated.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
#16332 fixed the missing
returnin DeepL's except branch, butComponentBase.be_outputwas removed during the agent refactor (#9113) while several components still call it. DeepL (and other tools) would raiseAttributeErrorbefore any error message could be returned.ComponentBase.be_outputaspd.DataFrame([{"content": v}])(same as pre-refactor behavior)_runreturns the**Error**:message when translation failsRelated to #16329
Test plan
test_run_returns_error_on_translation_failuretest_deepl.pycheck() tests still pass