remem-ai 0.5.96

Persistent memory for Claude Code and OpenAI Codex coding agents
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142

use std::io::{BufRead, BufReader};

use anyhow::Result;
use serde_json::Value;

use crate::ai::types::TokenUsage;

#[derive(Debug, Clone, PartialEq, Eq)]
pub(super) struct CodexRunUsage {
    pub usage: TokenUsage,
    pub model: Option<String>,
}

#[derive(Clone, Copy, Debug, Default, PartialEq, Eq)]
struct CodexCounters {
    input_tokens: i64,
    cached_input_tokens: i64,
    output_tokens: i64,
    reasoning_output_tokens: i64,
}

impl CodexCounters {
    fn from_value(value: &Value) -> Self {
        Self {
            input_tokens: json_i64(value, "input_tokens"),
            cached_input_tokens: json_i64(value, "cached_input_tokens")
                .max(json_i64(value, "cache_read_input_tokens")),
            output_tokens: json_i64(value, "output_tokens"),
            reasoning_output_tokens: json_i64(value, "reasoning_output_tokens"),
        }
    }

    fn is_empty(self) -> bool {
        self.input_tokens == 0
            && self.cached_input_tokens == 0
            && self.output_tokens == 0
            && self.reasoning_output_tokens == 0
    }

    fn into_token_usage(self) -> TokenUsage {
        let input_tokens = (self.input_tokens - self.cached_input_tokens).max(0);
        let output_tokens = (self.output_tokens - self.reasoning_output_tokens).max(0);
        TokenUsage {
            input_tokens,
            output_tokens,
            reasoning_tokens: self.reasoning_output_tokens,
            cache_read_tokens: self.cached_input_tokens,
            raw_input_tokens: self.input_tokens,
            raw_output_tokens: self.output_tokens,
            ..TokenUsage::default()
        }
    }
}

pub(super) fn parse_codex_json_events(
    events: &[u8],
    model: Option<String>,
) -> Result<Option<CodexRunUsage>> {
    let mut usage = TokenUsage::default();
    let reader = BufReader::new(events);

    for line in reader.lines() {
        let line = line?;
        let Ok(value) = serde_json::from_str::<Value>(&line) else {
            continue;
        };
        if value.get("type").and_then(Value::as_str) != Some("turn.completed") {
            continue;
        }
        let Some(raw_usage) = value.get("usage") else {
            continue;
        };
        let counters = CodexCounters::from_value(raw_usage);
        if counters.is_empty() {
            continue;
        }
        usage.add(&counters.into_token_usage());
    }

    if usage.is_empty() {
        Ok(None)
    } else {
        Ok(Some(CodexRunUsage { usage, model }))
    }
}

fn json_i64(value: &Value, key: &str) -> i64 {
    value.get(key).and_then(Value::as_i64).unwrap_or(0)
}

#[cfg(test)]
mod tests {
    use anyhow::{Context, Result};

    use super::parse_codex_json_events;

    #[test]
    fn parses_codex_exec_json_turn_completed_usage() -> Result<()> {
        let log = r#"
{"type":"thread.started","thread_id":"019e2f80-913f-7452-97ed-6340c56b2bd4"}
{"type":"turn.started"}
{"type":"item.completed","item":{"id":"item_0","type":"agent_message","text":"OK"}}
{"type":"turn.completed","usage":{"input_tokens":1000,"cached_input_tokens":200,"output_tokens":500,"reasoning_output_tokens":150}}
"#;
        let parsed = parse_codex_json_events(log.as_bytes(), Some("gpt-5.2".to_string()))
            .and_then(|usage| usage.context("usage should parse"))?;
        assert_eq!(parsed.model.as_deref(), Some("gpt-5.2"));
        assert_eq!(parsed.usage.input_tokens, 800);
        assert_eq!(parsed.usage.cache_read_tokens, 200);
        assert_eq!(parsed.usage.output_tokens, 350);
        assert_eq!(parsed.usage.reasoning_tokens, 150);
        assert_eq!(parsed.usage.total_tokens(), 1500);
        Ok(())
    }

    #[test]
    fn parses_multiple_codex_exec_json_turns() -> Result<()> {
        let log = r#"
{"type":"turn.completed","usage":{"input_tokens":100,"cached_input_tokens":80,"output_tokens":20,"reasoning_output_tokens":5}}
{"type":"turn.completed","usage":{"input_tokens":50,"cached_input_tokens":0,"output_tokens":10,"reasoning_output_tokens":0}}
"#;
        let parsed = parse_codex_json_events(log.as_bytes(), None)
            .and_then(|usage| usage.context("usage should parse"))?;
        assert_eq!(parsed.usage.input_tokens, 70);
        assert_eq!(parsed.usage.cache_read_tokens, 80);
        assert_eq!(parsed.usage.output_tokens, 25);
        assert_eq!(parsed.usage.reasoning_tokens, 5);
        assert_eq!(parsed.usage.total_tokens(), 180);
        Ok(())
    }

    #[test]
    fn returns_none_without_turn_completed_usage() -> Result<()> {
        let log = r#"
{"type":"thread.started","thread_id":"019e2f80-913f-7452-97ed-6340c56b2bd4"}
{"type":"item.completed","item":{"id":"item_0","type":"agent_message","text":"OK"}}
"#;
        let parsed = parse_codex_json_events(log.as_bytes(), Some("gpt-5.2".to_string()))?;
        assert!(parsed.is_none());
        Ok(())
    }
}