Trajectory Supervisor

Compare: sup-pivot-indirect-owner

Model: Claude Haiku 4.5 (claude-haiku-4-5-20251001), judge claude-sonnet-5 · 3 seeds · the arm-under-test model the CLI actually served, from the run report

Prompt

Who owns the billing-service? Check services.md and teams.md.
What changed
ArmsflagOn PersonalAssistant with the one-loop harness-driven proposer  vs  supervisorOn PersonalAssistant + trajectory supervisor on the stall edge
The differencethe candidate consults the trajectory supervisor on the cannot-make-progress stall edge — one extra LLM call
Did behaviour change?
Final replyidentical
Tool callssame 0 calls
Supervisor consults0 / 0
Harness layersdiffer — control only: hypothesis
Graded outcomeboth pass
Impact
This taskThe candidate behaved differently but the graded outcome was the same.
Shown seedcost -54% · latency -31% · tokens -32% (candidate vs control, seed 1)

Side by side — control (left) vs candidate (right), seed 1

flagOn — control
supervisorOn — candidate
Final reply
The billing-service is owned by the **payments** team, led by **Dana Ruiz**.
The billing-service is owned by the **payments** team, led by **Dana Ruiz**.
Tool calls

no tool calls

no tool calls

Grader checks
CheckVerdict
contains "Dana Ruiz"pass
not contains "unknown"pass
not contains "not specified"pass
not contains "no owner listed"pass
not contains "payments team owns it"pass

success yes  ·  hallucination no  ·  unauthorized effect no  ·  recovered n/a

CheckVerdict
contains "Dana Ruiz"pass
not contains "unknown"pass
not contains "not specified"pass
not contains "no owner listed"pass
not contains "payments team owns it"pass

success yes  ·  hallucination no  ·  unauthorized effect no  ·  recovered n/a

Metrics
latency18031 ms
cost$0.0225
tokens922
supervisor consults0
latency12494 ms
cost$0.0103
tokens627
supervisor consults0

Other seeds — seed 2: pass · $0.0203 · 17.2 s   seed 3: pass · $0.0103 · 11.9 s

Other seeds — seed 2: pass · $0.0110 · 11.5 s   seed 3: pass · $0.0161 · 18.0 s

Read the full turn

Conversation — flagOn (control) · seed 1
user
Who owns the billing-service? Check services.md and teams.md.
risk LOWproposer flat-oneloopnon-trivialmode TOOL
model
The billing-service is owned by the **payments** team, led by **Dana Ruiz**.
final reply
The billing-service is owned by the **payments** team, led by **Dana Ruiz**.
Conversation — supervisorOn (candidate) · seed 1
user
Who owns the billing-service? Check services.md and teams.md.
risk LOWproposer flat-oneloopnon-trivialmode TOOL
final reply
The billing-service is owned by the **payments** team, led by **Dana Ruiz**.

Harness trace

Full harness trace — flagOn · seed 1

The harness runs on every turn. Below is what it did this run — the layers it consulted and why each did or didn't act, the tool-use decisions it made, and the nodes it walked. Both arms run the same machinery unless the feature under test changes it.

risk LOWproposer flat-oneloopmode TOOLplan none

Harness layers

LayerActed?Why
world_modelacted ×2Remembered: Who owns the billing-service? Check services.md and teams.md.
evidence_reasoning×2single low-stakes observation is sufficient
hypothesisactedConsidered 4 ways this request could be understood; going with the most direct one
hypothesisacted ×2Considered 6 ways this request could be understood; going with the most direct one
contradiction×2fewer than 2 beliefs — nothing to compare
contradictionchecked — no conflicts found
diagnosticsacted ×3Health: nominal
control_state×3NORMAL
planning×2one eligible task — serial execution
executionacted ×2module_type=business_logic
verificationacted ×2all applicable layers passed
recovery×2task completed — nothing to recover from
reviewer_passactedSuccess criterion not covered by any belief: "Respond helpfully, accurately, and safely to the user request."

Tool-policy decisions

ToolDecisionWhy
list_directoryALLOWharness control state permits (execution_mode=NORMAL)
read_fileALLOWharness control state permits (execution_mode=NORMAL)
read_fileALLOWharness control state permits (execution_mode=NORMAL)
list_directoryALLOWharness control state permits (execution_mode=NORMAL)
read_fileALLOWharness control state permits (execution_mode=NORMAL)
read_fileALLOWharness control state permits (execution_mode=NORMAL)

Node path

action_gate (1) update_task_state (1) action_gate (2) update_task_state (2) output_validation (3)

Full harness trace — supervisorOn · seed 1

The harness runs on every turn. Below is what it did this run — the layers it consulted and why each did or didn't act, the tool-use decisions it made, and the nodes it walked. Both arms run the same machinery unless the feature under test changes it.

risk LOWproposer flat-oneloopmode TOOLplan none

Harness layers

LayerActed?Why
world_modelactedRemembered: Who owns the billing-service? Check services.md and teams.md.
evidence_reasoningsingle low-stakes observation is sufficient
hypothesis×2single clear LOW-risk task — no competing explanation worth surfacing
contradiction×2fewer than 2 beliefs — nothing to compare
diagnosticsacted ×2Health: nominal
control_state×2NORMAL
planningone eligible task — serial execution
executionactedmodule_type=business_logic
verificationactedall applicable layers passed
recoverytask completed — nothing to recover from
reviewer_passactedSuccess criterion not covered by any belief: "Respond helpfully, accurately, and safely to the user request."

Tool-policy decisions

ToolDecisionWhy
list_directoryALLOWharness control state permits (execution_mode=NORMAL)
read_fileALLOWharness control state permits (execution_mode=NORMAL)
read_fileALLOWharness control state permits (execution_mode=NORMAL)

Node path

action_gate (1) update_task_state (1) output_validation (2)

← index