omni-dev 0.44.0

AI-powered git commit rewriter, PR generator, and MCP server for Jira, Confluence, Datadog, Gmail, and Drive.
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

//! `omni-dev ai jev` — TypeSafe Jev "System One" typed judgments.
//!
//! Wraps [`crate::jev::client::JevClient`]'s single `system_one` call behind
//! four subcommands: `choice`, `score`, `noul` (single-question) and `ask`
//! (multi-question, from a file), plus `route`, which asks the stage questions
//! of [`crate::jev::route`] about GitHub issues, and `verify-decision`, which
//! checks a decision comment against the sources it cites
//! ([`crate::jev::verify`]). See `docs/jev.md` for the operator guide.

mod ask;
mod choice;
mod common;
mod noul;
mod route;
mod score;
mod verify_decision;

use anyhow::Result;
use clap::{Parser, Subcommand};

/// TypeSafe Jev typed judgments (choice, score, yes/no).
#[derive(Parser)]
#[command(
    long_about = "TypeSafe Jev typed judgments (choice, score, yes/no).\n\nJev is a \
separate typed-judgment API, not a chat model, so most jev subcommands do **not** accept \
the AI backend flags (`--ai-backend`, `--model`, `--claude-cli-*`, ...) — passing any \
of them is a clap error on those leaves. `route` and `verify-decision` are the two \
exceptions to `-C/--repo`, and `verify-decision` is the one subcommand that also accepts \
the AI backend flags, since it uses an AI backend to split a decision comment into \
statements before checking them with Jev. An exported `OMNI_DEV_MODEL` is silently ignored \
by every Jev call. Use `--jev-model` on each subcommand instead to override the Jev model."
)]
pub struct JevCommand {
    /// The jev subcommand to execute.
    #[command(subcommand)]
    pub command: JevSubcommands,
}

/// Jev subcommands.
#[derive(Subcommand)]
pub enum JevSubcommands {
    /// Picks one of a labelled set of options.
    Choice(choice::ChoiceCommand),
    /// Rates the state against an ordered scale.
    Score(score::ScoreCommand),
    /// Estimates the probability of a yes/no condition.
    Noul(noul::NoulCommand),
    /// Answers several questions about one state in a single pass, from a file.
    Ask(ask::AskCommand),
    /// Routes issues to model classes for their design, implement and review stages.
    Route(route::RouteCommand),
    /// Checks a decision comment against the issues or pull requests it cites.
    VerifyDecision(verify_decision::VerifyDecisionCommand),
}

impl JevCommand {
    /// Executes the jev command.
    pub async fn execute(self) -> Result<()> {
        match self.command {
            JevSubcommands::Choice(cmd) => cmd.execute().await,
            JevSubcommands::Score(cmd) => cmd.execute().await,
            JevSubcommands::Noul(cmd) => cmd.execute().await,
            JevSubcommands::Ask(cmd) => cmd.execute().await,
            JevSubcommands::Route(cmd) => cmd.execute().await,
            JevSubcommands::VerifyDecision(cmd) => cmd.execute().await,
        }
    }
}