anchor
stringlengths
723
2.15k
positive
stringlengths
153
8.2k
Before we proceed, can you verify that the currently available evidence is sufficient to resolve the contradiction about event delivery and retry behavior, without writing any state? Decision: route History: [{"action_family": "tool_use", "result": "weak", "step": 0}, {"action_family": "tool_use", "result": "weak", "st...
Apply this capability when the next decision depends on whether available information is sufficient.. Capabilities: assess evidence. Accepts: evidence, governance state. Returns: assessment. Evidence role: governance assessment. Prerequisites: requirements updated. Constraints: canonical evidence set required. Side eff...
Since we've already checked the metadata and found an acronym issue, what is the final step to complete the data transformation, given that all requirements are met? Decision: route History: [{"action_family": "tool_use", "result": "weak", "step": 0}] Plan: {"active": false, "operation": "enumerate_sources"} Agent Stat...
This candidate supports controlled inspection of the selected result or next action.. Capabilities: finalize selection. Accepts: assessment, governance state. Returns: terminal result. Evidence role: terminal decision. Prerequisites: requirements complete. Constraints: fresh sufficient assessment required. Side effects...
Should we search for the relevant technical instruction using a broad query, or is it better to first inspect the partially available context to narrow down the version noise? Decision: route Plan: {"active": true, "operation": "compatibility"} Agent State: {"phase": "retrieval", "question_length_band": "long", "state_...
This operation handles the best next information-gathering step while preserving the current task state.. Capabilities: plan retrieval. Accepts: agent state, text. Returns: plan. Evidence role: none. Prerequisites: objective available. Constraints: none. Side effects: local state write. Arguments: objective (string opt...
How can we verify the required data transformation by examining the available resources, given that we haven't inspected any evidence yet? Decision: route History: [{"action_family": "tool_use", "result": "weak", "step": 0}, {"action_family": "tool_use", "result": "weak", "step": 1}] Plan: {"active": false, "operation"...
This operation handles the available resources while preserving the current task state.. Capabilities: list sources. Accepts: metadata. Returns: source list. Evidence role: source discovery. Prerequisites: none. Constraints: none. Side effects: read. Arguments: filters (object optional), scope (string required)
Given limited steps left, which content should I prioritize to verify compatibility? Decision: route Plan: {"active": false, "operation": "compatibility"} Agent State: {"phase": "source_discovery", "question_length_band": "short", "state_name": "source_discovery"} Source State: {"available_modalities": ["excel", "metad...
Invoke the candidate to determine the relevant content from the available resources.. Capabilities: search content. Accepts: text, pdf. Returns: evidence candidates. Evidence role: candidate discovery. Prerequisites: query available. Constraints: indexed sources required. Side effects: read. Arguments: limit (integer o...
My previous attempt was incomplete; what exact identifier should I look up next? Decision: route History: [{"action_family": "tool_use", "result": "weak", "step": 0}, {"action_family": "tool_use", "result": "weak", "step": 1}, {"action_family": "tool_use", "result": "weak", "step": 2}, {"action_family": "tool_use", "re...
The operation produces a reliable result for an exact name, code, or token under current constraints.. Capabilities: exact pattern search. Accepts: text, code. Returns: evidence candidates. Evidence role: candidate discovery. Prerequisites: query available. Constraints: textual source required. Side effects: read. Argu...
What resource metadata and filters should I use to find the technical instruction? Decision: route Plan: {"active": false, "operation": "lookup_semantic"} Agent State: {"phase": "inspection", "question_length_band": "short", "state_name": "partial_evidence"} Source State: {"available_modalities": ["code", "metadata"], ...
Apply this capability when the next decision depends on resource metadata and filters.. Capabilities: search metadata. Accepts: metadata. Returns: source list. Evidence role: source discovery. Prerequisites: none. Constraints: metadata catalog required. Side effects: read. Arguments: filters (object required), sort (st...
What next action should I take to define fields and ensure compatibility? Decision: route History: [{"action_family": "tool_use", "result": "weak", "step": 0}, {"action_family": "tool_use", "result": "weak", "step": 1}, {"action_family": "tool_use", "result": "weak", "step": 2}, {"action_family": "tool_use", "result": ...
Use this candidate when the task requires fields, types, and structure.. Capabilities: inspect structured schema, discover fields. Accepts: csv, excel, sqlite. Returns: schema. Evidence role: structure discovery. Prerequisites: source selected. Constraints: structured source required. Side effects: read. Arguments: inc...
Find the matching records between the CSV files. Decision: route Plan: {"active": false, "operation": "lookup_exact"} Agent State: {"phase": "synthesis", "question_length_band": "short", "state_name": "contradiction"} Source State: {"available_modalities": ["csv", "metadata"], "evidence_topology_observed": "none_observ...
Use this candidate when the task requires matching records.. Capabilities: search structured records. Accepts: csv, excel, sqlite. Returns: evidence candidates. Evidence role: candidate discovery. Prerequisites: schema known. Constraints: structured source required. Side effects: read. Arguments: limit (integer require...
What are the section headings and page breaks in the PDF? Decision: route History: [{"action_family": "tool_use", "result": "weak", "step": 0}] Plan: {"active": false, "operation": "enumerate_records"} Agent State: {"phase": "terminal_readiness", "question_length_band": "short", "state_name": "fresh_sufficient"} Source...
This operation handles sections and structural landmarks while preserving the current task state.. Capabilities: inspect document structure. Accepts: pdf. Returns: document structure. Evidence role: structure discovery. Prerequisites: source selected. Constraints: pdf source required. Side effects: read. Arguments: lan...
Which pages discuss the version compatibility? Decision: route History: [{"action_family": "tool_use", "result": "weak", "step": 0}, {"action_family": "tool_use", "result": "weak", "step": 1}] Plan: {"active": false, "operation": "lookup_semantic"} Agent State: {"phase": "synthesis", "question_length_band": "short", "s...
Invoke the candidate to determine the relevant pages or ranges from the available resources.. Capabilities: search document pages. Accepts: pdf. Returns: evidence candidates. Evidence role: candidate discovery. Prerequisites: source selected. Constraints: pdf source required. Side effects: read. Arguments: page range (...
Compare the selected resource with the text sources for consistency. Decision: route History: [{"action_family": "tool_use", "result": "weak", "step": 0}, {"action_family": "tool_use", "result": "weak", "step": 1}, {"action_family": "tool_use", "result": "weak", "step": 2}, {"action_family": "tool_use", "result": "weak...
The operation produces a reliable result for the selected resource under current constraints.. Capabilities: read content. Accepts: text, code. Returns: content. Evidence role: evidence inspection. Prerequisites: source selected. Constraints: local source required. Side effects: read. Arguments: offset (integer optiona...
Given the governance phase and complete requirements, how should I proceed to resolve contradictions between sources, considering only read operations and avoiding state changes? Decision: route History: [{"action_family": "tool_use", "result": "weak", "step": 0}, {"action_family": "tool_use", "result": "weak", "step":...
Apply this capability when the next decision depends on symbols and definitions.. Capabilities: inspect code structure. Accepts: code. Returns: content, code structure. Evidence role: evidence inspection. Prerequisites: source selected. Constraints: code source required. Side effects: read. Arguments: depth (integer re...
We've checked some candidates but still have one unresolved requirement. What should we do next to identify the correct transformation, given the schema-bound query and semantic matching? Decision: route History: [{"action_family": "tool_use", "result": "weak", "step": 0}, {"action_family": "tool_use", "result": "weak"...
This candidate supports controlled inspection of a candidate result.. Capabilities: inspect evidence. Accepts: evidence candidates. Returns: evidence. Evidence role: evidence inspection. Prerequisites: evidence candidate selected. Constraints: candidate evidence required. Side effects: read. Arguments: claim (string op...
Between updating the stale assessment and retrieving the missing context for the technical instruction, which action is more appropriate for the next step in synthesis? Decision: route History: [{"action_family": "tool_use", "result": "weak", "step": 0}, {"action_family": "tool_use", "result": "weak", "step": 1}, {"act...
Use this candidate when the task requires missing surrounding context.. Capabilities: expand context. Accepts: evidence candidates, evidence. Returns: evidence. Evidence role: context expansion. Prerequisites: evidence candidate selected. Constraints: incomplete context required. Side effects: read. Arguments: result i...
How can we verify whether the two results agree or conflict, given that we've already inspected multiple sources and need to confirm absence? Decision: route History: [{"action_family": "tool_use", "result": "weak", "step": 0}, {"action_family": "tool_use", "result": "weak", "step": 1}] Plan: {"active": false, "operati...
This operation handles agreement or conflict between results while preserving the current task state.. Capabilities: compare evidence. Accepts: evidence. Returns: evidence comparison. Evidence role: evidence synthesis. Prerequisites: evidence inspected. Constraints: multiple evidence required. Side effects: none. Argum...
Given the remaining step budget, which next operation should I prioritize to verify the absence of the required evidence while maintaining task coverage? Decision: route Plan: {"active": true, "operation": "absence_verification"} Agent State: {"phase": "retrieval", "question_length_band": "long", "state_name": "retriev...
Invoke the candidate to determine task requirements and coverage from the available resources.. Capabilities: update requirements. Accepts: evidence, agent state. Returns: governance state. Evidence role: governance update. Prerequisites: evidence inspected. Constraints: tracked requirements required. Side effects: loc...
After the incomplete attempt, what should I do next to determine if the current information is sufficient for the access rules? Decision: route History: [{"action_family": "tool_use", "result": "weak", "step": 0}, {"action_family": "tool_use", "result": "weak", "step": 1}, {"action_family": "tool_use", "result": "weak"...
The operation produces a reliable result for whether available information is sufficient under current constraints.. Capabilities: assess evidence. Accepts: evidence, governance state. Returns: assessment. Evidence role: governance assessment. Prerequisites: requirements updated. Constraints: canonical evidence set req...
Which available source or catalog should I consult next to verify the absence of the event and decide on retry behavior? Decision: route History: [{"action_family": "tool_use", "result": "weak", "step": 0}, {"action_family": "tool_use", "result": "weak", "step": 1}, {"action_family": "tool_use", "result": "weak", "step...
Apply this capability when the next decision depends on the selected result or next action.. Capabilities: finalize selection. Accepts: assessment, governance state. Returns: terminal result. Evidence role: terminal decision. Prerequisites: requirements complete. Constraints: fresh sufficient assessment required. Side ...
What is the next action to gather information that resolves the contradiction about resource limits? Decision: route History: [{"action_family": "tool_use", "result": "weak", "step": 0}] Plan: {"active": true, "operation": "contradiction_resolution"} Agent State: {"phase": "planning", "question_length_band": "long", "s...
This candidate supports controlled inspection of the best next information-gathering step.. Capabilities: plan retrieval. Accepts: agent state, text. Returns: plan. Evidence role: none. Prerequisites: objective available. Constraints: none. Side effects: local state write. Arguments: objective (string required), priori...
We've inspected several sources and need to resolve a disputed requirement. What should we do next with the available resources? Decision: route History: [{"action_family": "tool_use", "result": "weak", "step": 0}] Plan: {"active": false, "operation": "compatibility"} Agent State: {"phase": "terminal_readiness", "quest...
Use this candidate when the task requires the available resources.. Capabilities: list sources. Accepts: metadata. Returns: source list. Evidence role: source discovery. Prerequisites: none. Constraints: none. Side effects: read. Arguments: filters (object optional), scope (string required)
We have snippets from a couple of sources and need to pick the latest value. What's the next step? Decision: route History: [{"action_family": "tool_use", "result": "weak", "step": 0}] Plan: {"active": false, "operation": "latest_value_selection"} Agent State: {"phase": "governance", "question_length_band": "short", "s...
This operation handles the relevant content while preserving the current task state.. Capabilities: search content. Accepts: text, pdf. Returns: evidence candidates. Evidence role: candidate discovery. Prerequisites: query available. Constraints: indexed sources required. Side effects: read. Arguments: limit (integer o...
We need to join complementary sources to perform the required transformation. What should we do next? Decision: route History: [{"action_family": "tool_use", "result": "weak", "step": 0}] Plan: {"active": false, "operation": "join"} Agent State: {"phase": "source_discovery", "question_length_band": "long", "state_name"...
The operation produces a reliable result for an exact name, code, or token under current constraints.. Capabilities: exact pattern search, identifier search. Accepts: text, code. Returns: evidence candidates. Evidence role: candidate discovery. Prerequisites: query available. Constraints: textual source required. Side ...
We need to resolve contradictions in the data. Given the unknown inventory and identifier variations, what should we do next? Decision: route Plan: {"active": false, "operation": "contradiction_resolution"} Agent State: {"phase": "terminal_readiness", "question_length_band": "long", "state_name": "fresh_sufficient"} So...
The operation produces a reliable result for resource metadata and filters under current constraints.. Capabilities: search metadata. Accepts: metadata. Returns: source list. Evidence role: source discovery. Prerequisites: none. Constraints: metadata catalog required. Side effects: read. Arguments: filters (object requ...
Given the stale schema assessment, which read-only operation should I perform next to get the field definitions and types? Decision: route History: [{"action_family": "tool_use", "result": "weak", "step": 0}, {"action_family": "tool_use", "result": "weak", "step": 1}, {"action_family": "tool_use", "result": "weak", "st...
Apply this capability when the next decision depends on fields, types, and structure.. Capabilities: inspect structured schema. Accepts: csv, excel, sqlite. Returns: schema. Evidence role: structure discovery. Prerequisites: source selected. Constraints: structured source required. Side effects: read. Arguments: includ...
Since we've checked the CSV records but not the lifecycle limits, what read-only step should I do next to match the remaining resources? Decision: route History: [{"action_family": "tool_use", "result": "weak", "step": 0}, {"action_family": "tool_use", "result": "weak", "step": 1}, {"action_family": "tool_use", "result...
This candidate supports controlled inspection of matching records.. Capabilities: search structured records. Accepts: csv, excel, sqlite. Returns: evidence candidates. Evidence role: candidate discovery. Prerequisites: schema known. Constraints: structured source required. Side effects: read. Arguments: limit (integer ...
Instead of reading the whole PDF, which read-only step should I take next to locate the relevant sections for validation? Decision: route History: [{"action_family": "tool_use", "result": "weak", "step": 0}, {"action_family": "tool_use", "result": "weak", "step": 1}, {"action_family": "tool_use", "result": "weak", "ste...
Use this candidate when the task requires sections and structural landmarks.. Capabilities: inspect document structure. Accepts: pdf. Returns: document structure. Evidence role: structure discovery. Prerequisites: source selected. Constraints: pdf source required. Side effects: read. Arguments: landmark (string optiona...
To verify the telemetry result, which read-only operation should I do next to inspect the relevant pages or ranges? Decision: route History: [{"action_family": "tool_use", "result": "weak", "step": 0}] Plan: {"active": false, "operation": "lookup_semantic"} Agent State: {"phase": "inspection", "question_length_band": "...
This operation handles the relevant pages or ranges while preserving the current task state.. Capabilities: search document pages. Accepts: pdf. Returns: evidence candidates. Evidence role: candidate discovery. Prerequisites: source selected. Constraints: pdf source required. Side effects: read. Arguments: page range (...
With only 8 steps left, which resource should I prioritize to join the split context and satisfy the remaining requirement? Decision: route Plan: {"active": false, "operation": "join"} Agent State: {"phase": "source_discovery", "question_length_band": "long", "state_name": "source_discovery"} Source State: {"available_...
Invoke the candidate to determine the selected resource from the available resources.. Capabilities: read content. Accepts: text, code. Returns: content. Evidence role: evidence inspection. Prerequisites: source selected. Constraints: local source required. Side effects: read. Arguments: offset (integer optional), reso...
My previous attempt was incomplete; what should I redo to resolve the contradiction involving the acronym and validate the quality constraints? Decision: route History: [{"action_family": "tool_use", "result": "weak", "step": 0}] Plan: {"active": false, "operation": "contradiction_resolution"} Agent State: {"phase": "i...
The operation produces a reliable result for symbols and definitions under current constraints.. Capabilities: inspect code structure. Accepts: code. Returns: content, code structure. Evidence role: evidence inspection. Prerequisites: source selected. Constraints: code source required. Side effects: read. Arguments: de...
Which available resource can provide the needed event delivery and retry behavior details for this join? Decision: route Plan: {"active": false, "operation": "join"} Agent State: {"phase": "governance", "question_length_band": "long", "state_name": "insufficient"} Source State: {"available_modalities": ["text", "pdf", ...
Apply this capability when the next decision depends on a candidate result.. Capabilities: inspect evidence. Accepts: evidence candidates. Returns: evidence. Evidence role: evidence inspection. Prerequisites: evidence candidate selected. Constraints: candidate evidence required. Side effects: read. Arguments: claim (st...
What should I do next to get the missing surrounding context for the exact identifier and resolve the contradiction? Decision: route History: [{"action_family": "tool_use", "result": "weak", "step": 0}, {"action_family": "tool_use", "result": "weak", "step": 1}] Plan: {"active": false, "operation": "contradiction_resol...
This candidate supports controlled inspection of missing surrounding context.. Capabilities: expand context. Accepts: evidence candidates, evidence. Returns: evidence. Evidence role: context expansion. Prerequisites: evidence candidate selected. Constraints: incomplete context required. Side effects: read. Arguments: r...
Compare the retrieved passages and tell me if they agree or conflict. Decision: route History: [{"action_family": "tool_use", "result": "weak", "step": 0}] Plan: {"active": true, "operation": "join"} Agent State: {"phase": "retrieval", "question_length_band": "short", "state_name": "retrieval"} Source State: {"availabl...
Use this candidate when the task requires agreement or conflict between results.. Capabilities: compare evidence. Accepts: evidence. Returns: evidence comparison. Evidence role: evidence synthesis. Prerequisites: evidence inspected. Constraints: multiple evidence required. Side effects: none. Arguments: comparison (str...
What should the next step be in the workflow? Decision: route History: [{"action_family": "tool_use", "result": "weak", "step": 0}, {"action_family": "tool_use", "result": "weak", "step": 1}, {"action_family": "tool_use", "result": "weak", "step": 2}, {"action_family": "tool_use", "result": "weak", "step": 3}] Plan: {"...
This operation handles task requirements and coverage while preserving the current task state.. Capabilities: update requirements. Accepts: evidence, agent state. Returns: governance state. Evidence role: governance update. Prerequisites: evidence inspected. Constraints: tracked requirements required. Side effects: loc...
Do we have enough information to deliver the event, or should we retry? Decision: route History: [{"action_family": "tool_use", "result": "weak", "step": 0}, {"action_family": "tool_use", "result": "weak", "step": 1}, {"action_family": "tool_use", "result": "weak", "step": 2}, {"action_family": "tool_use", "result": "w...
Invoke the candidate to determine whether available information is sufficient from the available resources.. Capabilities: assess evidence. Accepts: evidence, governance state. Returns: assessment. Evidence role: governance assessment. Prerequisites: requirements updated. Constraints: canonical evidence set required. S...
Based on the access rules and current state, what should we do next? Decision: route History: [{"action_family": "tool_use", "result": "weak", "step": 0}] Plan: {"active": true, "operation": "lookup_semantic"} Agent State: {"phase": "planning", "question_length_band": "short", "state_name": "initial"} Source State: {"a...
The operation produces a reliable result for the selected result or next action under current constraints.. Capabilities: finalize selection. Accepts: assessment, governance state. Returns: terminal result. Evidence role: terminal decision. Prerequisites: requirements complete. Constraints: fresh sufficient assessment ...
Given the stale access-control assessment and multiple inspected passages, what is the safest next step to enumerate records without writing local state? Decision: route History: [{"action_family": "tool_use", "result": "weak", "step": 0}] Plan: {"active": true, "operation": "enumerate_records"} Agent State: {"phase": ...
Apply this capability when the next decision depends on the best next information-gathering step.. Capabilities: plan retrieval. Accepts: agent state, text. Returns: plan. Evidence role: none. Prerequisites: objective available. Constraints: none. Side effects: local state write. Arguments: objective (string required),...
We've already checked several mixed sources but haven't inventoried them. What read-only operation should we perform next to discover available resources? Decision: route History: [{"action_family": "tool_use", "result": "weak", "step": 0}] Plan: {"active": true, "operation": "lookup_exact"} Agent State: {"phase": "ret...
This candidate supports controlled inspection of the available resources.. Capabilities: list sources. Accepts: metadata. Returns: source list. Evidence role: source discovery. Prerequisites: none. Constraints: none. Side effects: read. Arguments: filters (object optional), scope (string required)
Although we could inspect the code directly, what is the precise next step to retrieve relevant evidence about event delivery, given the evidence is fresh? Decision: route History: [{"action_family": "tool_use", "result": "weak", "step": 0}] Plan: {"active": true, "operation": "compare"} Agent State: {"phase": "retriev...
This operation handles the relevant content while preserving the current task state.. Capabilities: search content, retrieve passages. Accepts: text, pdf. Returns: evidence candidates. Evidence role: candidate discovery. Prerequisites: query available. Constraints: indexed sources required. Side effects: read. Argument...
How can we verify the exact event-handling code name by comparing complementary CSV sources, without modifying local state? Decision: route History: [{"action_family": "tool_use", "result": "weak", "step": 0}] Plan: {"active": false, "operation": "compatibility"} Agent State: {"phase": "inspection", "question_length_ba...
This operation handles an exact name, code, or token while preserving the current task state.. Capabilities: exact pattern search. Accepts: text, code. Returns: evidence candidates. Evidence role: candidate discovery. Prerequisites: query available. Constraints: textual source required. Side effects: read. Arguments: c...
Given the remaining step budget, prioritize which source metadata and filters to inspect next. Decision: route History: [{"action_family": "tool_use", "result": "weak", "step": 0}, {"action_family": "tool_use", "result": "weak", "step": 1}, {"action_family": "tool_use", "result": "weak", "step": 2}, {"action_family": "...
Invoke the candidate to determine resource metadata and filters from the available resources.. Capabilities: search metadata. Accepts: metadata. Returns: source list. Evidence role: source discovery. Prerequisites: none. Constraints: metadata catalog required. Side effects: read. Arguments: filters (object required), s...
The previous attempt was incomplete. What fields, types, and structure should we examine to proceed? Decision: route History: [{"action_family": "tool_use", "result": "weak", "step": 0}, {"action_family": "tool_use", "result": "weak", "step": 1}] Plan: {"active": false, "operation": "compatibility"} Agent State: {"phas...
The operation produces a reliable result for fields, types, and structure under current constraints.. Capabilities: inspect structured schema. Accepts: csv, excel, sqlite. Returns: schema. Evidence role: structure discovery. Prerequisites: source selected. Constraints: structured source required. Side effects: read. Ar...
Which records in the catalog match our criteria for failure recovery? Decision: route History: [{"action_family": "tool_use", "result": "weak", "step": 0}] Plan: {"active": false, "operation": "enumerate_records"} Agent State: {"phase": "governance", "question_length_band": "short", "state_name": "insufficient"} Source...
Apply this capability when the next decision depends on matching records.. Capabilities: search structured records. Accepts: csv, excel, sqlite. Returns: evidence candidates. Evidence role: candidate discovery. Prerequisites: schema known. Constraints: structured source required. Side effects: read. Arguments: limit (i...
What sections or structural landmarks should we look for to verify the absence? Decision: route History: [{"action_family": "tool_use", "result": "weak", "step": 0}, {"action_family": "tool_use", "result": "weak", "step": 1}] Plan: {"active": false, "operation": "absence_verification"} Agent State: {"phase": "inspectio...
Use this candidate when the task requires sections and structural landmarks.. Capabilities: inspect document structure, discover sections. Accepts: pdf. Returns: document structure. Evidence role: structure discovery. Prerequisites: source selected. Constraints: pdf source required. Side effects: read. Arguments: landm...
Find the relevant pages or ranges in the mixed sources to verify consistency. Decision: route History: [{"action_family": "tool_use", "result": "weak", "step": 0}] Plan: {"active": false, "operation": "latest_value_selection"} Agent State: {"phase": "source_discovery", "question_length_band": "short", "state_name": "so...
Use this candidate when the task requires the relevant pages or ranges.. Capabilities: search document pages. Accepts: pdf. Returns: evidence candidates. Evidence role: candidate discovery. Prerequisites: source selected. Constraints: pdf source required. Side effects: read. Arguments: page range (array required), quer...
List all records from the selected CSV resource for version compatibility. Decision: route History: [{"action_family": "tool_use", "result": "weak", "step": 0}, {"action_family": "tool_use", "result": "weak", "step": 1}] Plan: {"active": false, "operation": "enumerate_records"} Agent State: {"phase": "terminal_readines...
This operation handles the selected resource while preserving the current task state.. Capabilities: read content. Accepts: text, code. Returns: content. Evidence role: evidence inspection. Prerequisites: source selected. Constraints: local source required. Side effects: read. Arguments: offset (integer optional), reso...
Inspect the code symbols and definitions to plan event delivery and retry behavior. Decision: route Plan: {"active": false, "operation": "enumerate_records"} Agent State: {"phase": "governance", "question_length_band": "short", "state_name": "insufficient"} Source State: {"available_modalities": ["code", "metadata"], "...
Invoke the candidate to determine symbols and definitions from the available resources.. Capabilities: inspect code structure. Accepts: code. Returns: content, code structure. Evidence role: evidence inspection. Prerequisites: source selected. Constraints: code source required. Side effects: read. Arguments: depth (int...
Compare the candidate result with other evidence to assess consistency across sources. Decision: route History: [{"action_family": "tool_use", "result": "weak", "step": 0}] Plan: {"active": true, "operation": "compare"} Agent State: {"phase": "retrieval", "question_length_band": "short", "state_name": "retrieval"} Sour...
The operation produces a reliable result for a candidate result under current constraints.. Capabilities: inspect evidence. Accepts: evidence candidates. Returns: evidence. Evidence role: evidence inspection. Prerequisites: evidence candidate selected. Constraints: candidate evidence required. Side effects: read. Argum...
Compare the two evidence sources we found, but only use read-only operations and stick to structured data formats like CSV or SQLite, not PDFs. Decision: route Plan: {"active": false, "operation": "compare"} Agent State: {"phase": "governance", "question_length_band": "long", "state_name": "insufficient"} Source State:...
Apply this capability when the next decision depends on missing surrounding context.. Capabilities: expand context. Accepts: evidence candidates, evidence. Returns: evidence. Evidence role: context expansion. Prerequisites: evidence candidate selected. Constraints: incomplete context required. Side effects: read. Argum...
We've already inspected multiple sources and found schema differences. Given the stale assessment, what should we check next to resolve the field definition conflict? Decision: route History: [{"action_family": "tool_use", "result": "weak", "step": 0}] Plan: {"active": false, "operation": "compatibility"} Agent State: ...
This candidate supports controlled inspection of agreement or conflict between results.. Capabilities: compare evidence. Accepts: evidence. Returns: evidence comparison. Evidence role: evidence synthesis. Prerequisites: evidence inspected. Constraints: multiple evidence required. Side effects: none. Arguments: comparis...
We have full context but need to verify absence of a version identifier. Should we search for it in code or check governance state, rather than re-examining evidence? Decision: route History: [{"action_family": "tool_use", "result": "weak", "step": 0}] Plan: {"active": false, "operation": "absence_verification"} Agent ...
Use this candidate when the task requires task requirements and coverage.. Capabilities: update requirements. Accepts: evidence, agent state. Returns: governance state. Evidence role: governance update. Prerequisites: evidence inspected. Constraints: tracked requirements required. Side effects: local state write. Argum...
How can we verify that the next workflow step is compatible with the available PDF evidence? What read-only operation should we use to confirm sufficiency? Decision: route History: [{"action_family": "tool_use", "result": "weak", "step": 0}] Plan: {"active": true, "operation": "compatibility"} Agent State: {"phase": "r...
This operation handles whether available information is sufficient while preserving the current task state.. Capabilities: assess evidence. Accepts: evidence, governance state. Returns: assessment. Evidence role: governance assessment. Prerequisites: requirements updated. Constraints: canonical evidence set required. S...
With only two steps left, which action should I prioritize to enumerate records across these complementary sources, given their schemas differ? Decision: route Plan: {"active": true, "operation": "enumerate_records"} Agent State: {"phase": "planning", "question_length_band": "long", "state_name": "initial"} Source Stat...
Invoke the candidate to determine the selected result or next action from the available resources.. Capabilities: finalize selection. Accepts: assessment, governance state. Returns: terminal result. Evidence role: terminal decision. Prerequisites: requirements complete. Constraints: fresh sufficient assessment required...
My previous attempts failed due to version noise; what is the best next step to join the data from the known SQLite source, given I have one step left and two unresolved requirements? Decision: route History: [{"action_family": "tool_use", "result": "weak", "step": 0}, {"action_family": "tool_use", "result": "weak", "s...
The operation produces a reliable result for the best next information-gathering step under current constraints.. Capabilities: plan retrieval. Accepts: agent state, text. Returns: plan. Evidence role: none. Prerequisites: objective available. Constraints: none. Side effects: local state write. Arguments: objective (st...
I have no steps left and need to know what resources are available for joining across sources; can you list what I can access? Decision: route Plan: {"active": true, "operation": "join"} Agent State: {"phase": "retrieval", "question_length_band": "long", "state_name": "retrieval"} Source State: {"available_modalities":...
This candidate supports controlled inspection of the available resources.. Capabilities: list sources, discover sources. Accepts: metadata. Returns: source list. Evidence role: source discovery. Prerequisites: none. Constraints: none. Side effects: read. Arguments: filters (object required), scope (string optional)
What should I do next to retrieve the relevant content for validating data quality, considering the complementary sources and my current state? Decision: route History: [{"action_family": "tool_use", "result": "weak", "step": 0}, {"action_family": "tool_use", "result": "weak", "step": 1}] Plan: {"active": false, "opera...
This candidate supports controlled inspection of the relevant content.. Capabilities: search content. Accepts: text, pdf. Returns: evidence candidates. Evidence role: candidate discovery. Prerequisites: query available. Constraints: indexed sources required. Side effects: read. Arguments: limit (integer optional), quer...
Given the stale assessment and known source inventory, what is the exact access rule or permission token I should use next? Decision: route History: [{"action_family": "tool_use", "result": "weak", "step": 0}, {"action_family": "tool_use", "result": "weak", "step": 1}, {"action_family": "tool_use", "result": "weak", "s...
Use this candidate when the task requires an exact name, code, or token.. Capabilities: exact pattern search. Accepts: text, code. Returns: evidence candidates. Evidence role: candidate discovery. Prerequisites: query available. Constraints: textual source required. Side effects: read. Arguments: case sensitive (boolea...
What is the most recent value among the available resources, and what filters should I apply to identify it? Decision: route History: [{"action_family": "tool_use", "result": "weak", "step": 0}, {"action_family": "tool_use", "result": "weak", "step": 1}, {"action_family": "tool_use", "result": "weak", "step": 2}, {"act...
This operation handles resource metadata and filters while preserving the current task state.. Capabilities: search metadata. Accepts: metadata. Returns: source list. Evidence role: source discovery. Prerequisites: none. Constraints: metadata catalog required. Side effects: read. Arguments: filters (object required), s...
What fields, types, and structure should I use to combine the relevant passages from the multiple sources? Decision: route History: [{"action_family": "tool_use", "result": "weak", "step": 0}, {"action_family": "tool_use", "result": "weak", "step": 1}] Plan: {"active": false, "operation": "join"} Agent State: {"phase":...
Invoke the candidate to determine fields, types, and structure from the available resources.. Capabilities: inspect structured schema. Accepts: csv, excel, sqlite. Returns: schema. Evidence role: structure discovery. Prerequisites: source selected. Constraints: structured source required. Side effects: read. Arguments:...
Given the disputed progress and partial context, which matching records should I compare to resolve the contradiction in the CSV data? Decision: route History: [{"action_family": "tool_use", "result": "weak", "step": 0}, {"action_family": "tool_use", "result": "weak", "step": 1}] Plan: {"active": false, "operation": "c...
The operation produces a reliable result for matching records under current constraints.. Capabilities: search structured records. Accepts: csv, excel, sqlite. Returns: evidence candidates. Evidence role: candidate discovery. Prerequisites: schema known. Constraints: structured source required. Side effects: read. Argu...
Given that all requirements are complete and the assessment is stale, what read-only operation should we perform to map the document's structural landmarks and sections before finalizing readiness? Decision: route Plan: {"active": false, "operation": "enumerate_records"} Agent State: {"phase": "terminal_readiness", "qu...
Apply this capability when the next decision depends on sections and structural landmarks.. Capabilities: inspect document structure. Accepts: pdf. Returns: document structure. Evidence role: structure discovery. Prerequisites: source selected. Constraints: pdf source required. Side effects: read. Arguments: landmark (...
We've already searched twice with no inspection, and one requirement is unresolved. What stateful, read-only step should we take next to locate the relevant pages or ranges in the PDF for the entity-bound semantic lookup? Decision: route History: [{"action_family": "tool_use", "result": "weak", "step": 0}, {"action_fam...
This candidate supports controlled inspection of the relevant pages or ranges.. Capabilities: search document pages. Accepts: pdf. Returns: evidence candidates. Evidence role: candidate discovery. Prerequisites: source selected. Constraints: pdf source required. Side effects: read. Arguments: page range (array required...
We need to resolve a contradiction about resource limits. Instead of jumping to a content read, what read-only operation should we perform to identify the selected resource from the known Excel source inventory? Decision: route History: [{"action_family": "tool_use", "result": "weak", "step": 0}, {"action_family": "too...
Use this candidate when the task requires the selected resource.. Capabilities: read content. Accepts: text, code. Returns: content. Evidence role: evidence inspection. Prerequisites: source selected. Constraints: local source required. Side effects: read. Arguments: offset (integer optional), resource (string required...
After inspecting multiple sources and finding a disputed requirement, how can we verify the required data transformation by enumerating all sources that mention the acronym, using only read operations? Decision: route History: [{"action_family": "tool_use", "result": "weak", "step": 0}] Plan: {"active": true, "operatio...
This operation handles symbols and definitions while preserving the current task state.. Capabilities: inspect code structure. Accepts: code. Returns: content, code structure. Evidence role: evidence inspection. Prerequisites: source selected. Constraints: code source required. Side effects: read. Arguments: depth (int...
Given limited steps, which candidate result should I prioritize for validation and quality checks? Decision: route History: [{"action_family": "tool_use", "result": "weak", "step": 0}, {"action_family": "tool_use", "result": "weak", "step": 1}] Plan: {"active": false, "operation": "lookup_semantic"} Agent State: {"phas...
Invoke the candidate to determine a candidate result from the available resources.. Capabilities: inspect evidence. Accepts: evidence candidates. Returns: evidence. Evidence role: evidence inspection. Prerequisites: evidence candidate selected. Constraints: candidate evidence required. Side effects: read. Arguments: cl...
I need to recover from an incomplete attempt; what should I do next to get the missing surrounding context? Decision: route History: [{"action_family": "tool_use", "result": "weak", "step": 0}] Plan: {"active": false, "operation": "latest_value_selection"} Agent State: {"phase": "terminal_readiness", "question_length_b...
The operation produces a reliable result for missing surrounding context under current constraints.. Capabilities: expand context. Accepts: evidence candidates, evidence. Returns: evidence. Evidence role: context expansion. Prerequisites: evidence candidate selected. Constraints: incomplete context required. Side effec...
What resource or catalog should I consult to resolve the contradiction between results for the next workflow step? Decision: route History: [{"action_family": "tool_use", "result": "weak", "step": 0}] Plan: {"active": false, "operation": "contradiction_resolution"} Agent State: {"phase": "terminal_readiness", "question...
Apply this capability when the next decision depends on agreement or conflict between results.. Capabilities: compare evidence. Accepts: evidence. Returns: evidence comparison. Evidence role: evidence synthesis. Prerequisites: evidence inspected. Constraints: multiple evidence required. Side effects: none. Arguments: c...
What is the next action to cover all task requirements and advance the workflow? Decision: route Plan: {"active": false, "operation": "lookup_exact"} Agent State: {"phase": "governance", "question_length_band": "short", "state_name": "insufficient"} Source State: {"available_modalities": ["excel", "metadata"], "evidenc...
This candidate supports controlled inspection of task requirements and coverage.. Capabilities: update requirements. Accepts: evidence, agent state. Returns: governance state. Evidence role: governance update. Prerequisites: evidence inspected. Constraints: tracked requirements required. Side effects: local state write...
Can we proceed with what we have, or do we need more information? Decision: route History: [{"action_family": "tool_use", "result": "weak", "step": 0}, {"action_family": "tool_use", "result": "weak", "step": 1}] Plan: {"active": false, "operation": "enumerate_sources"} Agent State: {"phase": "synthesis", "question_leng...
Use this candidate when the task requires whether available information is sufficient.. Capabilities: assess evidence. Accepts: evidence, governance state. Returns: assessment. Evidence role: governance assessment. Prerequisites: requirements updated. Constraints: canonical evidence set required. Side effects: local st...
Which option should we pick next? Decision: route History: [{"action_family": "tool_use", "result": "weak", "step": 0}] Plan: {"active": false, "operation": "compatibility"} Agent State: {"phase": "governance", "question_length_band": "short", "state_name": "insufficient"} Source State: {"available_modalities": ["excel...
This operation handles the selected result or next action while preserving the current task state.. Capabilities: finalize selection. Accepts: assessment, governance state. Returns: terminal result. Evidence role: terminal decision. Prerequisites: requirements complete. Constraints: fresh sufficient assessment required...
What should we look for next to advance our goal? Decision: route History: [{"action_family": "tool_use", "result": "weak", "step": 0}, {"action_family": "tool_use", "result": "weak", "step": 1}] Plan: {"active": true, "operation": "enumerate_records"} Agent State: {"phase": "retrieval", "question_length_band": "short"...
Invoke the candidate to determine the best next information-gathering step from the available resources.. Capabilities: plan retrieval. Accepts: agent state, text. Returns: plan. Evidence role: none. Prerequisites: objective available. Constraints: none. Side effects: local state write. Arguments: objective (string req...
What resources are available to resolve this contradiction? Decision: route History: [{"action_family": "tool_use", "result": "weak", "step": 0}, {"action_family": "tool_use", "result": "weak", "step": 1}] Plan: {"active": false, "operation": "contradiction_resolution"} Agent State: {"phase": "source_discovery", "quest...
The operation produces a reliable result for the available resources under current constraints.. Capabilities: list sources. Accepts: metadata. Returns: source list. Evidence role: source discovery. Prerequisites: none. Constraints: none. Side effects: read. Arguments: filters (object optional), scope (string required)
Given the stale assessment and only snippets inspected, which next operation should I run to join complementary evidence from the PDF without writing local state? Decision: route History: [{"action_family": "tool_use", "result": "weak", "step": 0}] Plan: {"active": false, "operation": "join"} Agent State: {"phase": "go...
Apply this capability when the next decision depends on the relevant content.. Capabilities: search content. Accepts: text, pdf. Returns: evidence candidates. Evidence role: candidate discovery. Prerequisites: query available. Constraints: indexed sources required. Side effects: read. Arguments: limit (integer optional...
I've already checked some sources but haven't found the exact token. What remaining read-only operation should I perform to extract that precise identifier from the PDF, considering the planning phase? Decision: route Plan: {"active": true, "operation": "lookup_exact"} Agent State: {"phase": "planning", "question_lengt...
This candidate supports controlled inspection of an exact name, code, or token.. Capabilities: exact pattern search. Accepts: text, code. Returns: evidence candidates. Evidence role: candidate discovery. Prerequisites: query available. Constraints: textual source required. Side effects: read. Arguments: case sensitive ...
Instead of just listing all versions, what operation should I use to compare metadata and filter the Excel sources to find compatibility, updating local state as needed? Decision: route History: [{"action_family": "tool_use", "result": "weak", "step": 0}, {"action_family": "tool_use", "result": "weak", "step": 1}] Plan...
Use this candidate when the task requires resource metadata and filters.. Capabilities: search metadata. Accepts: metadata. Returns: source list. Evidence role: source discovery. Prerequisites: none. Constraints: metadata catalog required. Side effects: read. Arguments: filters (object required), sort (string optional)
How should I verify the disputed requirement by inspecting the CSV's fields and structure, using a read and local-state-write operation to gather exact evidence candidates? Decision: route History: [{"action_family": "tool_use", "result": "weak", "step": 0}] Plan: {"active": true, "operation": "lookup_exact"} Agent Sta...
This operation handles fields, types, and structure while preserving the current task state.. Capabilities: inspect structured schema. Accepts: csv, excel, sqlite. Returns: schema. Evidence role: structure discovery. Prerequisites: source selected. Constraints: structured source required. Side effects: read. Arguments:...
Given the disputed requirements and limited steps, which candidate best matches the records? Decision: route Plan: {"active": true, "operation": "enumerate_records"} Agent State: {"phase": "planning", "question_length_band": "short", "state_name": "initial"} Source State: {"available_modalities": ["csv", "metadata"], "...
Invoke the candidate to determine matching records from the available resources.. Capabilities: search structured records. Accepts: csv, excel, sqlite. Returns: evidence candidates. Evidence role: candidate discovery. Prerequisites: schema known. Constraints: structured source required. Side effects: read. Arguments: l...
After the incomplete attempt, what should be the next step to locate the relevant sections? Decision: route History: [{"action_family": "tool_use", "result": "weak", "step": 0}] Plan: {"active": false, "operation": "contradiction_resolution"} Agent State: {"phase": "governance", "question_length_band": "short", "state_...
The operation produces a reliable result for sections and structural landmarks under current constraints.. Capabilities: inspect document structure. Accepts: pdf. Returns: document structure. Evidence role: structure discovery. Prerequisites: source selected. Constraints: pdf source required. Side effects: read. Argume...
Which available resource can resolve the contradiction across sources, focusing on the relevant pages? Decision: route History: [{"action_family": "tool_use", "result": "weak", "step": 0}, {"action_family": "tool_use", "result": "weak", "step": 1}, {"action_family": "tool_use", "result": "weak", "step": 2}, {"action_fa...
Apply this capability when the next decision depends on the relevant pages or ranges.. Capabilities: search document pages. Accepts: pdf. Returns: evidence candidates. Evidence role: candidate discovery. Prerequisites: source selected. Constraints: pdf source required. Side effects: read. Arguments: page range (array r...
What is the next action to select the resource for telemetry and trace requirements? Decision: route History: [{"action_family": "tool_use", "result": "weak", "step": 0}] Plan: {"active": false, "operation": "latest_value_selection"} Agent State: {"phase": "inspection", "question_length_band": "short", "state_name": "p...
This candidate supports controlled inspection of the selected resource.. Capabilities: read content. Accepts: text, code. Returns: content. Evidence role: evidence inspection. Prerequisites: source selected. Constraints: local source required. Side effects: read. Arguments: offset (integer optional), resource (string r...
Given the stale assessment and unknown source inventory, what next operation should verify the absence of a specific identifier across mixed-format evidence? Decision: route History: [{"action_family": "tool_use", "result": "weak", "step": 0}] Plan: {"active": true, "operation": "absence_verification"} Agent State: {"p...
Use this candidate when the task requires symbols and definitions.. Capabilities: inspect code structure. Accepts: code. Returns: content, code structure. Evidence role: evidence inspection. Prerequisites: source selected. Constraints: code source required. Side effects: read. Arguments: depth (integer required), resou...
Considering the schema ambiguity and stale assessment, what should be done to enumerate records from the known sources, focusing on a candidate result? Decision: route History: [{"action_family": "tool_use", "result": "weak", "step": 0}] Plan: {"active": false, "operation": "enumerate_records"} Agent State: {"phase": "...
This operation handles a candidate result while preserving the current task state.. Capabilities: inspect evidence. Accepts: evidence candidates. Returns: evidence. Evidence role: evidence inspection. Prerequisites: evidence candidate selected. Constraints: candidate evidence required. Side effects: read. Arguments: cl...
To compare multiple passages from text sources and resolve identifier variation, what operation should retrieve the missing surrounding context? Decision: route History: [{"action_family": "tool_use", "result": "weak", "step": 0}, {"action_family": "tool_use", "result": "weak", "step": 1}] Plan: {"active": true, "opera...
Invoke the candidate to determine missing surrounding context from the available resources.. Capabilities: expand context. Accepts: evidence candidates, evidence. Returns: evidence. Evidence role: context expansion. Prerequisites: evidence candidate selected. Constraints: incomplete context required. Side effects: read...
Given the split context and complete requirements, what next operation should enumerate records to check agreement between results from mixed sources, considering the constraint of no side effects? Decision: route History: [{"action_family": "tool_use", "result": "weak", "step": 0}, {"action_family": "tool_use", "resul...
The operation produces a reliable result for agreement or conflict between results under current constraints.. Capabilities: compare evidence. Accepts: evidence. Returns: evidence comparison. Evidence role: evidence synthesis. Prerequisites: evidence inspected. Constraints: multiple evidence required. Side effects: non...
We have partial context from 8 sources and need to enumerate all sources, but avoid any side effects beyond local writes and reads. What should we do next? Decision: route History: [{"action_family": "tool_use", "result": "weak", "step": 0}] Plan: {"active": false, "operation": "enumerate_sources"} Agent State: {"phase...
Apply this capability when the next decision depends on task requirements and coverage.. Capabilities: update requirements. Accepts: evidence, agent state. Returns: governance state. Evidence role: governance update. Prerequisites: evidence inspected. Constraints: tracked requirements required. Side effects: local stat...
We've checked two sources and found nothing, but the schema is unclear. What operation should we run next to verify absence without missing anything? Decision: route History: [{"action_family": "tool_use", "result": "weak", "step": 0}, {"action_family": "tool_use", "result": "weak", "step": 1}] Plan: {"active": true, "...
This candidate supports controlled inspection of whether available information is sufficient.. Capabilities: assess evidence. Accepts: evidence, governance state. Returns: assessment. Evidence role: governance assessment. Prerequisites: requirements updated. Constraints: canonical evidence set required. Side effects: l...
We have full context and fresh assessment, but need the exact next action. Should we look up the identifier or just pick a candidate from the list? Decision: route Plan: {"active": false, "operation": "lookup_semantic"} Agent State: {"phase": "synthesis", "question_length_band": "long", "state_name": "contradiction"} S...
Use this candidate when the task requires the selected result or next action.. Capabilities: finalize selection. Accepts: assessment, governance state. Returns: terminal result. Evidence role: terminal decision. Prerequisites: requirements complete. Constraints: fresh sufficient assessment required. Side effects: local...
We only have snippets and need the latest value, but schema is ambiguous. What information-gathering step should we take to verify the result? Decision: route History: [{"action_family": "tool_use", "result": "weak", "step": 0}, {"action_family": "tool_use", "result": "weak", "step": 1}, {"action_family": "tool_use", "...
This operation handles the best next information-gathering step while preserving the current task state.. Capabilities: plan retrieval. Accepts: agent state, text. Returns: plan. Evidence role: none. Prerequisites: objective available. Constraints: none. Side effects: local state write. Arguments: objective (string req...
Given only 2 steps left and a stale assessment, should I prioritize inspecting the available resources for field definitions and compatibility? Decision: route History: [{"action_family": "tool_use", "result": "weak", "step": 0}, {"action_family": "tool_use", "result": "weak", "step": 1}] Plan: {"active": false, "opera...
Invoke the candidate to determine the available resources from the available resources.. Capabilities: list sources. Accepts: metadata. Returns: source list. Evidence role: source discovery. Prerequisites: none. Constraints: none. Side effects: read. Arguments: filters (object optional), scope (string required)
The previous attempt was incomplete; how should I recover by focusing on the relevant content for failure handling and recovery? Decision: route Plan: {"active": true, "operation": "compare"} Agent State: {"phase": "planning", "question_length_band": "long", "state_name": "initial"} Source State: {"available_modalities...
The operation produces a reliable result for the relevant content under current constraints.. Capabilities: search content. Accepts: text, pdf. Returns: evidence candidates. Evidence role: candidate discovery. Prerequisites: query available. Constraints: indexed sources required. Side effects: read. Arguments: limit (i...
Which available resource provides the exact name, code, or token for field definitions and compatibility? Decision: route Plan: {"active": false, "operation": "latest_value_selection"} Agent State: {"phase": "synthesis", "question_length_band": "long", "state_name": "contradiction"} Source State: {"available_modalities...
This candidate supports controlled inspection of an exact name, code, or token.. Capabilities: exact pattern search, identifier search. Accepts: text, code. Returns: evidence candidates. Evidence role: candidate discovery. Prerequisites: query available. Constraints: textual source required. Side effects: read. Argumen...
What should I do next to retrieve resource metadata and filters for the upcoming workflow step? Decision: route History: [{"action_family": "tool_use", "result": "weak", "step": 0}, {"action_family": "tool_use", "result": "weak", "step": 1}, {"action_family": "tool_use", "result": "weak", "step": 2}, {"action_family": ...
This candidate supports controlled inspection of resource metadata and filters.. Capabilities: search metadata. Accepts: metadata. Returns: source list. Evidence role: source discovery. Prerequisites: none. Constraints: metadata catalog required. Side effects: read. Arguments: filters (object required), sort (string op...
Compare the field definitions and types across the two candidate schemas to identify structural differences. Decision: route History: [{"action_family": "tool_use", "result": "weak", "step": 0}, {"action_family": "tool_use", "result": "weak", "step": 1}] Plan: {"active": false, "operation": "compare"} Agent State: {"ph...
Use this candidate when the task requires fields, types, and structure.. Capabilities: inspect structured schema. Accepts: csv, excel, sqlite. Returns: schema. Evidence role: structure discovery. Prerequisites: source selected. Constraints: structured source required. Side effects: read. Arguments: include types (boole...
Which candidate record aligns with the relevant technical instruction? Decision: route History: [{"action_family": "tool_use", "result": "weak", "step": 0}] Plan: {"active": false, "operation": "compatibility"} Agent State: {"phase": "governance", "question_length_band": "short", "state_name": "insufficient"} Source St...
This operation handles matching records while preserving the current task state.. Capabilities: search structured records. Accepts: csv, excel, sqlite. Returns: evidence candidates. Evidence role: candidate discovery. Prerequisites: schema known. Constraints: structured source required. Side effects: read. Arguments: l...
To proceed with the next workflow step, identify the section and structural landmarks in the source that indicate the join point. Decision: route History: [{"action_family": "tool_use", "result": "weak", "step": 0}, {"action_family": "tool_use", "result": "weak", "step": 1}, {"action_family": "tool_use", "result": "wea...
Invoke the candidate to determine sections and structural landmarks from the available resources.. Capabilities: inspect document structure. Accepts: pdf. Returns: document structure. Evidence role: structure discovery. Prerequisites: source selected. Constraints: pdf source required. Side effects: read. Arguments: lan...
Given the disputed field definitions, find the pages or ranges in the evidence that resolve the contradiction. Decision: route History: [{"action_family": "tool_use", "result": "weak", "step": 0}, {"action_family": "tool_use", "result": "weak", "step": 1}] Plan: {"active": true, "operation": "contradiction_resolution"}...
The operation produces a reliable result for the relevant pages or ranges under current constraints.. Capabilities: search document pages. Accepts: pdf. Returns: evidence candidates. Evidence role: candidate discovery. Prerequisites: source selected. Constraints: pdf source required. Side effects: read. Arguments: page...
Before finalizing, what read-only operation should we run on the selected resource to verify its content against the stale evidence? Decision: route History: [{"action_family": "tool_use", "result": "weak", "step": 0}, {"action_family": "tool_use", "result": "weak", "step": 1}, {"action_family": "tool_use", "result": "...
Apply this capability when the next decision depends on the selected resource.. Capabilities: read content. Accepts: text, code. Returns: content. Evidence role: evidence inspection. Prerequisites: source selected. Constraints: local source required. Side effects: read. Arguments: offset (integer optional), resource (s...
We've only seen snippets so far; what should we consult next to compare the multiple passages and resolve the acronym ambiguity? Decision: route Plan: {"active": false, "operation": "compare"} Agent State: {"phase": "synthesis", "question_length_band": "long", "state_name": "contradiction"} Source State: {"available_mo...
This candidate supports controlled inspection of symbols and definitions.. Capabilities: inspect code structure. Accepts: code. Returns: content, code structure. Evidence role: evidence inspection. Prerequisites: source selected. Constraints: code source required. Side effects: read. Arguments: depth (integer required)...
Instead of enumerating all sources, what specific action should we take on the candidate result to check version compatibility? Decision: route History: [{"action_family": "tool_use", "result": "weak", "step": 0}, {"action_family": "tool_use", "result": "weak", "step": 1}] Plan: {"active": false, "operation": "enumerat...
Use this candidate when the task requires a candidate result.. Capabilities: inspect evidence. Accepts: evidence candidates. Returns: evidence. Evidence role: evidence inspection. Prerequisites: evidence candidate selected. Constraints: candidate evidence required. Side effects: read. Arguments: claim (string optional)...
To verify the exact identifier's access rules, what evidence should we retrieve that includes the missing surrounding context? Decision: route Plan: {"active": false, "operation": "lookup_exact"} Agent State: {"phase": "synthesis", "question_length_band": "long", "state_name": "contradiction"} Source State: {"available...
This operation handles missing surrounding context while preserving the current task state.. Capabilities: expand context. Accepts: evidence candidates, evidence. Returns: evidence. Evidence role: context expansion. Prerequisites: evidence candidate selected. Constraints: incomplete context required. Side effects: read...