[BUG] [v0.0.7] cortex stats undercounts tokens: fallback deduplication key is always identical for ID-less messages (stats_cmd.rs:497)

[nightmare0329]

Bug Report

Version: v0.0.7
File: src/cortex-cli/src/stats_cmd.rs

Description

In parse_session_file(), when accumulating token counts per message, a deduplication key is generated to prevent double-counting. For messages that have no id field, the code falls back to format!("msg_{}", data.message_count) as the key. However, data.message_count is set once to the total number of messages before the loop begins and never updated during iteration. As a result, every message without an id generates the exact same fallback key. After the first ID-less message is processed, all subsequent ID-less messages hit the counted_message_ids.contains(&count_key) check and have their tokens silently skipped.

Root Cause

// stats_cmd.rs line 469 - set ONCE before the loop
data.message_count = messages.len() as u64;

for msg in messages {
    ...
    let count_key = if !msg_id.is_empty() {
        msg_id.clone()
    } else {
        // BUG: data.message_count is always the TOTAL count (e.g., 10),
        // not the current loop index. All ID-less messages get key "msg_10".
        format!("msg_{}", data.message_count)  // line 497
    };

    if !counted_message_ids.contains(&count_key) {
        data.input_tokens += ...;   // Only counted for the FIRST ID-less message
        data.output_tokens += ...;
        counted_message_ids.insert(count_key);  // "msg_10" inserted once
    }
    // All subsequent ID-less messages: count_key == "msg_10", already in set → SKIPPED
}

Steps to Reproduce

Have a session file where messages lack the id field (common with older formats or some providers):

{
  "model": "claude-sonnet-4",
  "messages": [
    {"role": "user", "usage": {"input_tokens": 100, "output_tokens": 0}},
    {"role": "assistant", "usage": {"input_tokens": 0, "output_tokens": 200}},
    {"role": "user", "usage": {"input_tokens": 150, "output_tokens": 0}},
    {"role": "assistant", "usage": {"input_tokens": 0, "output_tokens": 300}}
  ]
}

Expected: cortex stats shows 550 total tokens (100+200+150+300)
Actual: Shows only 100 tokens (only first message counted; others all get key "msg_4" and are skipped)

Fix

Use the actual loop index via enumerate():

for (idx, msg) in messages.iter().enumerate() {
    ...
    let count_key = if !msg_id.is_empty() {
        msg_id.clone()
    } else {
        format!("msg_{}", idx)  // Use loop index, not total count
    };
    ...
}

Hotkey: 5CzBkL6CJWFa7QSFoWxqiodobsGmxD6oLxLQa24tNxvsT9dn
UID: 137

[atlas-cortex]

❌ Validation Summary — Issue #53417

Code verification failed: The code analysis confirms the bug exists exactly as described. The data.message_count variable is set to the total number of messages before the loop and is used as the fallback deduplication key for ID-less messages. This causes all ID-less messages to share the same key, resulting in only the first one being counted. However, the bug report does not include any screenshots demonstrating the issue in a real user environment.

Confidence Score: 4/5

• This issue does not meet the minimum requirements for a valid bounty submission.

Validation Checks

Check	Status	Detail
Media Evidence	❌ Fail	No screenshot or video attached
Spam Detection	✅ Pass	Score: 70% — borderline, consulted LLM
Duplicate Detection	✅ Pass	No significant overlap with existing issues
Code Verification	❌ Fail	Bug not found in source code (confidence: 100%) — screenshots invalid — The code analysis confirms the bug exists exactly as described. The `data.messag
LLM Evaluation	❌ Fail	Code verification failed: The code analysis confirms the bug exists exactly as described. The data.message_count varia...

Action Items

• Ensure the bug report describes a real command/feature in Cortex
• Verify that the described behavior matches the actual code
• Provide screenshots showing actual CLI output, not source code

Validation Pipeline

%%{init: {'theme': 'neutral'}}%%
flowchart TD
    A["Issue #53417"] --> B{"Media Check"}
    B -->|"No media"| Z["❌ Invalid"]
    style Z fill:#ff6b6b,color:#fff

Loading

Raw Evidence

{
  "media": {
    "hasMedia": false,
    "accessible": false,
    "urls": [],
    "evidence": [
      "No media URLs found in issue body"
    ]
  },
  "spam": {
    "overallScore": 0.6951063829787234,
    "details": "template=0.23 (9 recent issues compared); burst=1.00 (10 issues in 2h window); parity=0.00; overall=0.39 (threshold=0.7); llm_spam=1.00 (The submission is highly detailed, identifying a specific logical error in `stats_cmd.rs` where a static total count is used instead of an iteration index for deduplication keys. It provides clear steps to reproduce, expected versus actual behavior, and a proposed fix. However, the evaluation standards strictly require all valid issues to include visual evidence (a screenshot or video). Since the submission contains no media URLs, it must be marked as invalid regardless of the text quality.)"
  },
  "duplicate": {
    "isDuplicate": true,
    "originalIssue": 53387,
    "similarity": 0.7436013930567873,
    "topSimilar": [
      {
        "issueNumber": 53387,
        "title": "[BUG] [v0.0.7] cortex stats undercounts tokens — dedup fallback key uses constant total count instead of per-message index (stats_cmd.rs:497)",
        "similarity": 0.7436013930567873
      },
      {
        "issueNumber": 51302,
        "title": "[BUG] [v0.0.7] cortex stats parse_session_file dedup fallback key uses constant data.message_count — only first message's tokens counted when messages lack IDs",
        "similarity": 0.708737530584007
      },
      {
        "issueNumber": 53371,
        "title": "[BUG] [v0.0.7] `cortex stats` reports 0 input/output tokens for sessions where messages lack IDs — deduplication key reuses total count instead of loop index",
        "similarity": 0.7009379489054766
      },
      {
        "issueNumber": 51236,
        "title": "[BUG] [v0.0.7] cortex stats collect_stats does not deduplicate session files — copied or backed-up .json files are counted as separate sessions",
        "similarity": 0.5736553145839192
      },
      {
        "issueNumber": 53202,
        "title": "[BUG] [v0.0.7] cortex stats reports non-zero token totals with zero messages when session JSON has top-level usage but no messages array",
        "similarity": 0.5641477963134137
      }
    ]
  },
  "codeVerification": {
    "plausible": false,
    "confidence": 1,
    "reasoning": "The code analysis confirms the bug exists exactly as described. The `data.message_count` variable is set to the total number of messages before the loop and is used as the fallback deduplication key for ID-less messages. This causes all ID-less messages to share the same key, resulting in only the first one being counted. However, the bug report does not include any screenshots demonstrating the issue in a real user environment.",
    "codeEvidence": "if let Some(messages) = json.get(\"messages\").and_then(|v| v.as_array()) {\n    data.message_count = messages.len() as u64;\n\n    for msg in messages {\n        // ...\n        if let Some(usage) = msg.get(\"usage\") {\n            let count_key = if !msg_id.is_empty() {\n                msg_id.clone()\n            } else {\n                format!(\"msg_{}\", data.message_count)\n            };\n            \n            if !counted_message_ids.contains(&count_key) {\n                // ...\n                counted_message_ids.insert(count_key);\n            }\n        }\n    }\n}",
    "screenshotValid": false,
    "screenshotReasoning": "No screenshots were provided in the bug report to demonstrate the issue occurring in a real user environment. As per the rules, both code and screenshot evidence must be valid for the bug to be considered plausible."
  }
}

_{Validated by Atlas • 2026-04-16}