paperclip-runner evalsstatic evidence

← Overview

Agent blocks its assigned task

block-task

The agent durably blocks the task on the known catalog dependency.

Prompt

Block your assigned task because the semantic catalog must be republished. Set task-blocker as the blocking task.

Fixture

../fixtures/companies/orientation.json

Authority

{
  "actorId": "actor-engineer",
  "capabilities": [],
  "explicitClaims": [],
  "taskId": "task-orientation"
}

Current results

ModelConfigurationResultEstimated costRun time
claude-sonnet-5live-acpx-0.13.1-claude-acp-0.70.0-sonnet-5PASS$0.067297 estimated2026-09-06T21:22:22.764Z
claude-sonnet-5live-claude-managed-agents-2026-04-01-sonnet-5PASS$0.013331 estimated2026-09-06T21:22:28.478Z
global.anthropic.claude-sonnet-4-6live-aws-agentcore-global-anthropic-claude-sonnet-4-6PASS$0.039165 estimated2026-09-06T21:21:54.529Z
gpt-5.4-minilive-codex-pinnedPASS$0.001316 estimated2026-09-06T21:24:03.405Z
gpt-5.6-sollive-acpx-0.13.1-codex-acp-1.6.2-controlPASS$0.009400 estimated2026-09-06T21:22:13.741Z
openrouter/anthropic/claude-sonnet-5live-opencode-1.18.17-openrouter-claude-sonnet-5PASS$0.021816 estimated2026-09-06T21:24:25.907Z
openrouter/deepseek/deepseek-v4-flash-0731live-opencode-1.18.17-openrouter-deepseek-v4-flash-0731PASS$0.000950 estimated2026-09-06T21:28:20.332Z
openrouter/google/gemini-3.8-flashlive-opencode-1.18.17-openrouter-gemini-3.8-flashPASS$0.024607 estimated2026-09-06T21:24:47.030Z
openrouter/openai/gpt-6-astralive-opencode-1.18.17-openrouter-gpt-6-astraPASS$0.038663 estimated2026-09-06T21:26:05.096Z
openrouter/qwen/qwen3.8-max-0902live-opencode-1.18.17-openrouter-qwen3.8-max-0902PASS$0.016830 estimated2026-09-06T21:27:21.071Z
openrouter/z-ai/glm-5.3live-opencode-1.18.17-openrouter-glm-5.3PASS$0.014809 estimated2026-09-06T21:28:46.451Z

Assertions

block
semantic_operation
{
  "id": "block",
  "kind": "semantic_operation",
  "maxCalls": 1,
  "minCalls": 1,
  "operationId": "block_task",
  "requireSuccess": true
}
task-blocked
state_record_field
{
  "collection": "tasks",
  "equals": "blocked",
  "field": "status",
  "id": "task-blocked",
  "kind": "state_record_field",
  "recordId": "task-orientation"
}
terminal-completed
terminal
{
  "id": "terminal-completed",
  "kind": "terminal",
  "status": "completed"
}
transcript-complete
transcript_complete
{
  "id": "transcript-complete",
  "kind": "transcript_complete"
}
mock-only
network
{
  "id": "mock-only",
  "kind": "network",
  "realPaperclipRequests": 0
}
Full authored test definition
{
  "authority": {
    "actorId": "actor-engineer",
    "capabilities": [],
    "explicitClaims": [],
    "taskId": "task-orientation"
  },
  "checks": [
    {
      "id": "block",
      "kind": "semantic_operation",
      "maxCalls": 1,
      "minCalls": 1,
      "operationId": "block_task",
      "requireSuccess": true
    },
    {
      "collection": "tasks",
      "equals": "blocked",
      "field": "status",
      "id": "task-blocked",
      "kind": "state_record_field",
      "recordId": "task-orientation"
    },
    {
      "id": "terminal-completed",
      "kind": "terminal",
      "status": "completed"
    },
    {
      "id": "transcript-complete",
      "kind": "transcript_complete"
    },
    {
      "id": "mock-only",
      "kind": "network",
      "realPaperclipRequests": 0
    }
  ],
  "description": "The agent durably blocks the task on the known catalog dependency.",
  "fixture": "../fixtures/companies/orientation.json",
  "id": "block-task",
  "prompt": "Block your assigned task because the semantic catalog must be republished. Set task-blocker as the blocking task.",
  "schema": "paperclip-runner/eval-case/v1",
  "title": "Agent blocks its assigned task"
}