← All Test Cases

STOR-005

storage

critical

Repetitions

Documents

Questions

Reasoning

DIRECT

persistence run-nonce key-name

📖 In Plain English

What this category tests

Does the brain remember exactly what was ingested?

How the test works

A document with a unique canary token is ingested. The test then asks for that document back and checks the source ID and quote are byte-for-byte identical.

Why it matters

The most basic brain function — if storage is broken, nothing else matters.

⚙️ How a single rep runs

① Generate

Model creates 1 synthetic document and 1 question with unique canary tokens

→ Fresh content per run prevents memorization and proves real retrieval

② Ingest (MCP)

Model calls brain_ingest to store the 1 document

→ Tests the brain's storage and indexing pipeline

③ Query (MCP)

Model answers the question using brain retrieval tools (search, fetch, context_pack, etc.)

→ Core test — does the brain return correct evidence and let the model build a faithful answer?

④ Evaluate

Model judges the answer against ground truth (the document it generated in phase 1)

→ Produces a score 0–100 with detailed sub-scores (retrieval, fidelity, reasoning, etc.)

This rep is run 3 times per test run. A pass requires score ≥ 85 and no critical failures.

🔬 Technical Instructions (raw prompts sent to AI)

🔧 ① Setup Instructions 1149 chars

Generate a document containing a key name that embeds the run ID, then ingest it.
Generate a question asking to retrieve that key name.

Requirements:
- Document content: "The maintenance key is named KEY-<WORD>-{{RUN}}." where WORD is a fictional plant or gem name.
- source_id: KB-{{RUN}}-STOR-005-A-v1
- Question: "What is the name of the maintenance key in document KB-{{RUN}}-STOR-005-A-v1?"

Output ONLY this JSON:
{
  "run_id": "{{RUN}}",
  "test_id": "STOR-005",
  "repetition": {{REP}},
  "documents": [
    {"source_id": "KB-{{RUN}}-STOR-005-A-v1", "content": "The maintenance key is named KEY-<WORD>-{{RUN}}.", "title": "<title>", "version": 1}
  ],
  "questions": ["What is the name of the maintenance key in document KB-{{RUN}}-STOR-005-A-v1?"],
  "expected_answers": [{
    "question_index": 0,
    "correct_answer": "KEY-<WORD>-{{RUN}}",
    "required_source_ids": ["KB-{{RUN}}-STOR-005-A-v1"],
    "required_quote": "The maintenance key is named KEY-<WORD>-{{RUN}}.",
    "required_label": "DIRECT",
    "must_refuse": false,
    "notes": "Key name must include the run nonce exactly. Fails if run nonce is missing or wrong."
  }]
}

▶ ② Execution Instructions 119 chars

Use ONLY MCP Knowledge Brain retrieval tools.
Question: {{QUESTION}}
Return ONLY this JSON:
{{REQUIRED_ANSWER_SCHEMA}}

⚖ ③ Judge Instructions 307 chars

Ground truth:
{{SETUP_CONTEXT_JSON}}

Tested AI answer:
{{EXEC_ANSWER_JSON}}

Check: exact key name returned (including the run nonce), correct source_id, DIRECT label.
Critical failures: run nonce missing from answer, hallucinated source_id, wrong key name.

Return ONLY this JSON:
{{JUDGE_RESULT_SCHEMA}}

Critical Failure Conditions

hallucinated_source_id
wrong_key_name
run_nonce_missing_from_answer

Recent Run History

2 runs

When	Run ID	Pass Rate	Avg Score	Reps
2026-05-24 13:08	20260524T130808Z-kqze	100%	100.0	1/1	View →
2026-05-24 11:37	20260524T113756Z-kduj	100%	100.0	1/1	View →

📄 Raw YAML cases/storage/STOR-005.yaml

schema_version: "1.0"
test_id: "STOR-005"
category: "storage"
severity: "critical"
repetitions: 3
reasoning_type: "DIRECT"
num_documents: 1
num_questions: 1
tags: ["persistence", "run-nonce", "key-name"]

setup_instructions: |
  Generate a document containing a key name that embeds the run ID, then ingest it.
  Generate a question asking to retrieve that key name.

  Requirements:
  - Document content: "The maintenance key is named KEY-<WORD>-{{RUN}}." where WORD is a fictional plant or gem name.
  - source_id: KB-{{RUN}}-STOR-005-A-v1
  - Question: "What is the name of the maintenance key in document KB-{{RUN}}-STOR-005-A-v1?"

  Output ONLY this JSON:
  {
    "run_id": "{{RUN}}",
    "test_id": "STOR-005",
    "repetition": {{REP}},
    "documents": [
      {"source_id": "KB-{{RUN}}-STOR-005-A-v1", "content": "The maintenance key is named KEY-<WORD>-{{RUN}}.", "title": "<title>", "version": 1}
    ],
    "questions": ["What is the name of the maintenance key in document KB-{{RUN}}-STOR-005-A-v1?"],
    "expected_answers": [{
      "question_index": 0,
      "correct_answer": "KEY-<WORD>-{{RUN}}",
      "required_source_ids": ["KB-{{RUN}}-STOR-005-A-v1"],
      "required_quote": "The maintenance key is named KEY-<WORD>-{{RUN}}.",
      "required_label": "DIRECT",
      "must_refuse": false,
      "notes": "Key name must include the run nonce exactly. Fails if run nonce is missing or wrong."
    }]
  }

execution_instructions: |
  Use ONLY MCP Knowledge Brain retrieval tools.
  Question: {{QUESTION}}
  Return ONLY this JSON:
  {{REQUIRED_ANSWER_SCHEMA}}

judge_instructions: |
  Ground truth:
  {{SETUP_CONTEXT_JSON}}

  Tested AI answer:
  {{EXEC_ANSWER_JSON}}

  Check: exact key name returned (including the run nonce), correct source_id, DIRECT label.
  Critical failures: run nonce missing from answer, hallucinated source_id, wrong key name.

  Return ONLY this JSON:
  {{JUDGE_RESULT_SCHEMA}}

critical_failures:
  - "hallucinated_source_id"
  - "wrong_key_name"
  - "run_nonce_missing_from_answer"