chnjinlei/hive
1
0
Fork 0
Code Issues Pull Requests Actions 39 Packages Projects Releases Wiki Activity

including missing tool log inputs

Browse Source
This commit is contained in:
bryan
2026-02-05 09:08:42 -08:00
parent 7519c73f2a
commit 305b880b1d
3 changed files with 54 additions and 27 deletions
Download Patch File Download Diff File Expand all files Collapse all files
.gitignore
+1
View File
@@ -74,3 +74,4 @@ exports/*
docs/github-issues/*
core/tests/*dumps/*
screenshots/*
core/framework/graph/event_loop_node.py
+52 -27
View File
@@ -305,6 +305,7 @@ class EventLoopNode(NodeProtocol):
real_tool_results,
outputs_set,
turn_tokens,
logged_tool_calls,
) = await self._run_single_turn(ctx, conversation, tools, iteration, accumulator)
logger.info(
"[%s] iter=%d: LLM done — text=%d chars, real_tools=%d, "
@@ -370,7 +371,7 @@ class EventLoopNode(NodeProtocol):
step_index=iteration,
verdict="CONTINUE",
verdict_feedback="Stall detected before judge evaluation",
tool_calls=real_tool_results,
tool_calls=logged_tool_calls,
llm_text=assistant_text,
input_tokens=turn_tokens.get("input", 0),
output_tokens=turn_tokens.get("output", 0),
@@ -430,7 +431,7 @@ class EventLoopNode(NodeProtocol):
step_index=iteration,
verdict="CONTINUE",
verdict_feedback="Shutdown signaled (client-facing)",
tool_calls=real_tool_results,
tool_calls=logged_tool_calls,
llm_text=assistant_text,
input_tokens=turn_tokens.get("input", 0),
output_tokens=turn_tokens.get("output", 0),
@@ -474,7 +475,7 @@ class EventLoopNode(NodeProtocol):
step_index=iteration,
verdict="CONTINUE",
verdict_feedback="No input received (shutdown during wait)",
tool_calls=real_tool_results,
tool_calls=logged_tool_calls,
llm_text=assistant_text,
input_tokens=turn_tokens.get("input", 0),
output_tokens=turn_tokens.get("output", 0),
@@ -524,7 +525,7 @@ class EventLoopNode(NodeProtocol):
step_index=iteration,
verdict="CONTINUE",
verdict_feedback="Unjudged (judge_every_n_turns skip)",
tool_calls=real_tool_results,
tool_calls=logged_tool_calls,
llm_text=assistant_text,
input_tokens=turn_tokens.get("input", 0),
output_tokens=turn_tokens.get("output", 0),
@@ -579,7 +580,7 @@ class EventLoopNode(NodeProtocol):
verdict_feedback=(
f"Judge accepted but missing output keys: {missing}"
),
tool_calls=real_tool_results,
tool_calls=logged_tool_calls,
llm_text=assistant_text,
input_tokens=turn_tokens.get("input", 0),
output_tokens=turn_tokens.get("output", 0),
@@ -603,7 +604,7 @@ class EventLoopNode(NodeProtocol):
step_index=iteration,
verdict="ACCEPT",
verdict_feedback=verdict.feedback,
tool_calls=real_tool_results,
tool_calls=logged_tool_calls,
llm_text=assistant_text,
input_tokens=turn_tokens.get("input", 0),
output_tokens=turn_tokens.get("output", 0),
@@ -645,7 +646,7 @@ class EventLoopNode(NodeProtocol):
step_index=iteration,
verdict="ESCALATE",
verdict_feedback=verdict.feedback,
tool_calls=real_tool_results,
tool_calls=logged_tool_calls,
llm_text=assistant_text,
input_tokens=turn_tokens.get("input", 0),
output_tokens=turn_tokens.get("output", 0),
@@ -686,7 +687,7 @@ class EventLoopNode(NodeProtocol):
step_index=iteration,
verdict="RETRY",
verdict_feedback=verdict.feedback,
tool_calls=real_tool_results,
tool_calls=logged_tool_calls,
llm_text=assistant_text,
input_tokens=turn_tokens.get("input", 0),
output_tokens=turn_tokens.get("output", 0),
@@ -774,16 +775,21 @@ class EventLoopNode(NodeProtocol):
tools: list[Tool],
iteration: int,
accumulator: OutputAccumulator,
) -> tuple[str, list[dict], list[str], dict[str, int]]:
) -> tuple[str, list[dict], list[str], dict[str, int], list[dict]]:
"""Run a single LLM turn with streaming and tool execution.
Returns (assistant_text, real_tool_results, outputs_set, token_counts).
Returns (assistant_text, real_tool_results, outputs_set, token_counts, logged_tool_calls).
``real_tool_results`` contains only results from actual tools (web_search,
etc.), NOT from the synthetic ``set_output`` tool. ``outputs_set`` lists
the output keys written via ``set_output`` during this turn. This
separation lets the caller treat set_output as a framework concern
rather than a tool-execution concern.
``logged_tool_calls`` accumulates ALL tool calls across inner iterations
(real tools, set_output, and discarded calls) for L3 logging. Unlike
``real_tool_results`` which resets each inner iteration, this list grows
across the entire turn.
"""
stream_id = ctx.node_id
node_id = ctx.node_id
@@ -792,6 +798,9 @@ class EventLoopNode(NodeProtocol):
final_text = ""
# Track output keys set via set_output across all inner iterations
outputs_set_this_turn: list[str] = []
# Accumulate ALL tool calls across inner iterations for L3 logging.
# Unlike real_tool_results (reset each inner iteration), this persists.
logged_tool_calls: list[dict] = []
# Inner tool loop: stream may produce tool calls requiring re-invocation
while True:
@@ -862,7 +871,7 @@ class EventLoopNode(NodeProtocol):
# If no tool calls, turn is complete
if not tool_calls:
return final_text, [], outputs_set_this_turn, token_counts
return final_text, [], outputs_set_this_turn, token_counts, logged_tool_calls
# Execute tool calls — separate real tools from set_output
real_tool_results: list[dict] = []
@@ -909,18 +918,28 @@ class EventLoopNode(NodeProtocol):
pass
await accumulator.set(tc.tool_input["key"], value)
outputs_set_this_turn.append(tc.tool_input["key"])
else:
# --- Real tool execution ---
result = await self._execute_tool(tc)
result = self._truncate_tool_result(result, tc.tool_name)
real_tool_results.append(
logged_tool_calls.append(
{
"tool_use_id": tc.tool_use_id,
"tool_name": tc.tool_name,
"tool_name": "set_output",
"tool_input": tc.tool_input,
"content": result.content,
"is_error": result.is_error,
}
)
else:
# --- Real tool execution ---
result = await self._execute_tool(tc)
result = self._truncate_tool_result(result, tc.tool_name)
tool_entry = {
"tool_use_id": tc.tool_use_id,
"tool_name": tc.tool_name,
"tool_input": tc.tool_input,
"content": result.content,
"is_error": result.is_error,
}
real_tool_results.append(tool_entry)
logged_tool_calls.append(tool_entry)
# Record tool result in conversation (both real and set_output
# go into the conversation for LLM context continuity)
@@ -967,15 +986,15 @@ class EventLoopNode(NodeProtocol):
)
# Discarded calls go into real_tool_results so the
# caller sees they were attempted (for judge context).
real_tool_results.append(
{
"tool_use_id": tc.tool_use_id,
"tool_name": tc.tool_name,
"tool_input": tc.tool_input,
"content": discard_msg,
"is_error": True,
}
)
discard_entry = {
"tool_use_id": tc.tool_use_id,
"tool_name": tc.tool_name,
"tool_input": tc.tool_input,
"content": discard_msg,
"is_error": True,
}
real_tool_results.append(discard_entry)
logged_tool_calls.append(discard_entry)
# Prune old tool results NOW to prevent context bloat on the
# next turn. The char-based token estimator underestimates
# actual API tokens, so the standard compaction check in the
@@ -993,7 +1012,13 @@ class EventLoopNode(NodeProtocol):
)
# Limit hit — return from this turn so the judge can
# evaluate instead of looping back for another stream.
return final_text, real_tool_results, outputs_set_this_turn, token_counts
return (
final_text,
real_tool_results,
outputs_set_this_turn,
token_counts,
logged_tool_calls,
)
# --- Mid-turn pruning: prevent context blowup within a single turn ---
if conversation.usage_ratio() >= 0.6:
core/framework/runtime/agent_runtime.py
+1
View File
@@ -468,6 +468,7 @@ def create_agent_runtime(
tools: Available tools
tool_executor: Tool executor function
config: Runtime configuration
runtime_log_store: Optional RuntimeLogStore for per-execution logging
Returns:
Configured AgentRuntime (not yet started)