{% extends "_base.html" %} {% block title %}harness-evaluator — {{ run_name }} — {{ cell.cell_id }}{% endblock %} {% block content %}

{{ cell.cell_id }}

{% set exit_class = 'pass' if cell.exit_class == 'pass' else ('partial' if cell.exit_class == 'partial' else ('kill' if cell.exit_class in ('retryable_kill', 'non_retryable_kill') else 'fail')) %}
{{ cell.exit_class }}
Exit Class
{{ "%.2f"|format(cell.success) }}
Success
${{ "%.6f"|format(cell.total_cost) }}
Cost
{{ "%.1f"|format(cell.latency_ms / 1000) }}s
Latency

Cell Details

Run
{{ cell.run_name }}
Harness
{{ cell.harness }}
Model
{{ cell.model }}
Task
{{ cell.task_id }}
Track
{{ cell.track }}
Repeat
r{{ cell.repeat }}
Timestamp
{{ cell.timestamp }}
{% if cell.retry_count %}
Retries
{{ cell.retry_count }}
{% endif %}

Token Usage

InputOutputCache ReadCache WriteReasoningTotal
{{ cell.input_tokens }} {{ cell.output_tokens }} {{ cell.cache_read_tokens }} {{ cell.cache_write_tokens }} {{ cell.reasoning_tokens }} {{ cell.input_tokens + cell.output_tokens + cell.cache_read_tokens + cell.cache_write_tokens + cell.reasoning_tokens }}
API CallsTool CallsTime to First Attempt
{{ cell.num_api_calls }} {{ cell.num_tool_calls }} {{ "%.1f"|format(cell.time_to_first_attempt_ms / 1000) }}s
{% if cell.error_class or cell.error_message %}

Error

{% if cell.error_class %}

Error Class: {{ cell.error_class }}

{% endif %} {% if cell.error_message %}
{{ cell.error_message }}
{% endif %}
{% endif %} {% if cell.harness_stdout or cell.harness_stderr %}

Harness Output

{% if cell.harness_stderr %}
stderr ({{ cell.harness_stderr|length }} chars)
{{ cell.harness_stderr }}
{% endif %} {% if cell.harness_stdout %}
stdout ({{ cell.harness_stdout|length }} chars)
{{ cell.harness_stdout }}
{% endif %}
{% else %}

Harness Output

No harness output was captured for this cell.

{% endif %} {% if cell.diff %}

Diff

{{ cell.diff }}
{% endif %} {% if cell.test_output %}

Test Output

{{ cell.test_output }}
{% endif %} {% if cell.harness_metadata %}

Harness Metadata

{{ cell.harness_metadata }}
{% endif %} {% if phases %}

Phase Results

{% for p in phases %} {% endfor %}
PhaseModelRole Exit CodeDurationInput Tokens Output TokensCostAPI Calls Timed Out
{{ p.phase_name }} {{ p.model }} {{ p.model_role }} {{ p.exit_code }} {{ "%.1f"|format(p.duration_ms / 1000) }}s {{ p.input_tokens }} {{ p.output_tokens }} ${{ "%.6f"|format(p.total_cost) }} {{ p.num_api_calls }} {{ "yes" if p.timed_out else "no" }}
{% for p in phases %} {% if p.stdout or p.stderr %}
{{ p.phase_name }} output {% if p.stderr %}

stderr:

{{ p.stderr }}
{% endif %} {% if p.stdout %}

stdout:

{{ p.stdout }}
{% endif %}
{% endif %} {% endfor %}
{% endif %} {% if reconciliation %}

Reconciliation

Matched: {{ "yes" if reconciliation.matched else "no" }}

Max Discrepancy: {{ "%.1f"|format(reconciliation.max_discrepancy_pct) }}%

{% if reconciliation.proxy_usage_json %}

Proxy Usage:

{{ reconciliation.proxy_usage_json }}
{% endif %} {% if reconciliation.self_reported_usage_json %}

Self-Reported Usage:

{{ reconciliation.self_reported_usage_json }}
{% endif %} {% if reconciliation.details_json %}

Details:

{{ reconciliation.details_json }}
{% endif %}
{% endif %} {% endblock %}