fix(agent): detect truncated streaming tool calls before execution#6776
Closed
AIandI0x1 wants to merge 1 commit into
Closed
fix(agent): detect truncated streaming tool calls before execution#6776AIandI0x1 wants to merge 1 commit into
AIandI0x1 wants to merge 1 commit into
Conversation
teknium1
pushed a commit
that referenced
this pull request
Apr 9, 2026
When a streaming response is cut mid-tool-call (connection drop, timeout), the accumulated function.arguments is invalid JSON. The mock response builder defaulted finish_reason to 'stop', so the agent loop treated it as a valid completed turn and tried to execute tools with broken args. Fix: validate tool call arguments with json.loads() during mock response reconstruction. If any are invalid JSON, override finish_reason to 'length'. In the main loop's length handler, if tool calls are present, refuse to execute and return partial=True with a clear error instead of silently failing or wasting retries. Also fixes _thinking_exhausted to not short-circuit when tool calls are present — truncated tool calls are not thinking exhaustion. Original cherry-picked from PR #6776 by AIandI0x1. Closes #6638.
teknium1
pushed a commit
that referenced
this pull request
Apr 10, 2026
When a streaming response is cut mid-tool-call (connection drop, timeout), the accumulated function.arguments is invalid JSON. The mock response builder defaulted finish_reason to 'stop', so the agent loop treated it as a valid completed turn and tried to execute tools with broken args. Fix: validate tool call arguments with json.loads() during mock response reconstruction. If any are invalid JSON, override finish_reason to 'length'. In the main loop's length handler, if tool calls are present, refuse to execute and return partial=True with a clear error instead of silently failing or wasting retries. Also fixes _thinking_exhausted to not short-circuit when tool calls are present — truncated tool calls are not thinking exhaustion. Original cherry-picked from PR #6776 by AIandI0x1. Closes #6638.
Contributor
|
Merged via PR #6847. Your commit was cherry-picked onto current main with authorship preserved. We added a single retry before refusing — if the first attempt has truncated tool args, the agent retries the API call once before giving up. Thanks for the fix! |
Tommyeds
pushed a commit
to Tommyeds/hermes-agent
that referenced
this pull request
Apr 12, 2026
When a streaming response is cut mid-tool-call (connection drop, timeout), the accumulated function.arguments is invalid JSON. The mock response builder defaulted finish_reason to 'stop', so the agent loop treated it as a valid completed turn and tried to execute tools with broken args. Fix: validate tool call arguments with json.loads() during mock response reconstruction. If any are invalid JSON, override finish_reason to 'length'. In the main loop's length handler, if tool calls are present, refuse to execute and return partial=True with a clear error instead of silently failing or wasting retries. Also fixes _thinking_exhausted to not short-circuit when tool calls are present — truncated tool calls are not thinking exhaustion. Original cherry-picked from PR NousResearch#6776 by AIandI0x1. Closes NousResearch#6638.
3 tasks
ulasbilgen
pushed a commit
to ulasbilgen/hermes-adhd-agent
that referenced
this pull request
May 1, 2026
When a streaming response is cut mid-tool-call (connection drop, timeout), the accumulated function.arguments is invalid JSON. The mock response builder defaulted finish_reason to 'stop', so the agent loop treated it as a valid completed turn and tried to execute tools with broken args. Fix: validate tool call arguments with json.loads() during mock response reconstruction. If any are invalid JSON, override finish_reason to 'length'. In the main loop's length handler, if tool calls are present, refuse to execute and return partial=True with a clear error instead of silently failing or wasting retries. Also fixes _thinking_exhausted to not short-circuit when tool calls are present — truncated tool calls are not thinking exhaustion. Original cherry-picked from PR NousResearch#6776 by AIandI0x1. Closes NousResearch#6638.
02356abc
pushed a commit
to 02356abc/hermes-agent
that referenced
this pull request
May 14, 2026
When a streaming response is cut mid-tool-call (connection drop, timeout), the accumulated function.arguments is invalid JSON. The mock response builder defaulted finish_reason to 'stop', so the agent loop treated it as a valid completed turn and tried to execute tools with broken args. Fix: validate tool call arguments with json.loads() during mock response reconstruction. If any are invalid JSON, override finish_reason to 'length'. In the main loop's length handler, if tool calls are present, refuse to execute and return partial=True with a clear error instead of silently failing or wasting retries. Also fixes _thinking_exhausted to not short-circuit when tool calls are present — truncated tool calls are not thinking exhaustion. Original cherry-picked from PR NousResearch#6776 by AIandI0x1. Closes NousResearch#6638.
olympus-terminal
pushed a commit
to olympus-terminal/hermes-agent
that referenced
this pull request
May 16, 2026
When a streaming response is cut mid-tool-call (connection drop, timeout), the accumulated function.arguments is invalid JSON. The mock response builder defaulted finish_reason to 'stop', so the agent loop treated it as a valid completed turn and tried to execute tools with broken args. Fix: validate tool call arguments with json.loads() during mock response reconstruction. If any are invalid JSON, override finish_reason to 'length'. In the main loop's length handler, if tool calls are present, refuse to execute and return partial=True with a clear error instead of silently failing or wasting retries. Also fixes _thinking_exhausted to not short-circuit when tool calls are present — truncated tool calls are not thinking exhaustion. Original cherry-picked from PR NousResearch#6776 by AIandI0x1. Closes NousResearch#6638.
gweeteve
pushed a commit
to gweeteve/hermes-agent
that referenced
this pull request
Jun 2, 2026
When a streaming response is cut mid-tool-call (connection drop, timeout), the accumulated function.arguments is invalid JSON. The mock response builder defaulted finish_reason to 'stop', so the agent loop treated it as a valid completed turn and tried to execute tools with broken args. Fix: validate tool call arguments with json.loads() during mock response reconstruction. If any are invalid JSON, override finish_reason to 'length'. In the main loop's length handler, if tool calls are present, refuse to execute and return partial=True with a clear error instead of silently failing or wasting retries. Also fixes _thinking_exhausted to not short-circuit when tool calls are present — truncated tool calls are not thinking exhaustion. Original cherry-picked from PR NousResearch#6776 by AIandI0x1. Closes NousResearch#6638.
Egavasyug
pushed a commit
to Egavasyug/hermes-agent
that referenced
this pull request
Jun 10, 2026
When a streaming response is cut mid-tool-call (connection drop, timeout), the accumulated function.arguments is invalid JSON. The mock response builder defaulted finish_reason to 'stop', so the agent loop treated it as a valid completed turn and tried to execute tools with broken args. Fix: validate tool call arguments with json.loads() during mock response reconstruction. If any are invalid JSON, override finish_reason to 'length'. In the main loop's length handler, if tool calls are present, refuse to execute and return partial=True with a clear error instead of silently failing or wasting retries. Also fixes _thinking_exhausted to not short-circuit when tool calls are present — truncated tool calls are not thinking exhaustion. Original cherry-picked from PR NousResearch#6776 by AIandI0x1. Closes NousResearch#6638.
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
2 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
finish_reason="length"run_conversation()behaviorRoot cause
The OpenAI-compatible streaming response builder reconstructed a mock chat-completions response and defaulted missing finish reasons to
"stop".If the stream died mid-tool-call, the accumulated
function.argumentscould be invalid/incomplete JSON, but the rebuilt response still looked like a normal completed assistant turn.That meant truncated tool-call responses could be misclassified as successful completions instead of surfaced as truncation failures.
Fix
"length"lengthresponse contains tool calls, return a partial truncation error immediately and do not execute any toolsValidation
python -m pytest tests/run_agent/test_run_agent.py::TestStreamingApiCall::test_truncated_tool_call_args_upgrade_finish_reason_to_length tests/run_agent/test_run_agent.py::TestRunConversation::test_length_with_tool_calls_returns_partial_without_executing_tools tests/run_agent/test_run_agent.py::TestRunConversation::test_length_finish_reason_requests_continuation -qpython -m pytest tests/run_agent/test_run_agent.py -qpython -m py_compile run_agent.py tests/run_agent/test_run_agent.pyAll of the above passed locally.
Closes #6638