ContribAI Codebase Summary

Version: 5.2.0 | Language: Rust | Total LOC: ~22,000 | Files: 63 .rs | Tests: 335

Quick Navigation

crates/contribai-rs/src/
├── core/              # Config, models, middleware, events, errors, quotas (9 files)
├── llm/               # Multi-provider LLM routing, context, formatter (7 files)
├── github/            # GitHub REST+GraphQL client, discovery, guidelines (4 files)
├── analysis/          # Analyzers, skills, triage, AST intel, repo map (8 files)
├── generator/         # Code fix generation, scorer, self-review, validation (6 files)
├── orchestrator/      # Pipeline, memory (SQLite), review gate (4 files)
├── pr/                # PR lifecycle, patrol monitoring (3 files)
├── issues/            # Issue-driven contributions (2 files)
├── agents/            # Sub-agent registry (2 files)
├── tools/             # Tool protocol (1 file)
├── mcp/               # MCP server (21 tools) + client (3 files)
├── web/               # Axum dashboard, API key auth, webhooks (1 file)
├── cli/               # Clap CLI with 22 commands + ratatui TUI (3 files)
├── scheduler/         # Tokio cron scheduler (1 file)
├── plugins/           # Trait-based plugin system (1 file)
├── templates/         # Contribution templates (1 file)
├── notifications/     # Slack/Discord/Telegram (1 file)
├── sandbox/           # Code validation (1 file)
├── lib.rs             # Crate root (public modules)
└── main.rs            # Binary entry point

Module Responsibilities

Module	Purpose	Key Types/Functions	Files
core	Config, models, middleware, events, errors, quotas, profiles, retry	`ContribAIConfig`, `Repository`, `Finding`, `Contribution`, `Middleware`, `EventBus`	9
llm	Multi-provider routing, token budgeting, formatting	`LlmProvider`, `TaskRouter`, `ContextManager`, `Formatter`	7
github	Async GitHub client (REST+GraphQL), repo discovery, guideline parsing	`GitHubClient`, `RepoDiscovery`, `GuidelineParser`	4
analysis	Multi-strategy analysis, progressive skills, triage, AST intel, repo map	`CodeAnalyzer`, `AnalysisSkill`, `TriageEngine`, `AstIntel`, `RepoMap`	8
generator	LLM-powered fix generation, self-review, quality scoring, validation	`ContributionGenerator`, `QualityScorer`, `SelfReview`	6
orchestrator	Pipeline coordination, SQLite memory, review gate	`Pipeline`, `Memory`, `ReviewGate`	4
pr	PR creation, patrol monitoring	`PRManager`, `PRPatrol`	3
issues	Issue discovery and solving	`IssueSolver`	2
agents	Sub-agent registry with parallel execution	`SubAgentRegistry`	2
tools	Tool protocol (MCP-inspired)	`Tool`, `ToolResult`	1
mcp	MCP stdio server (21 tools) + client	`McpServer`, `StdioMcpClient`	3
web	Axum REST API, API key auth, webhook receiver, dashboard	`run_server`, `AppState`, `verify_webhook_signature`	1
cli	Clap-based CLI with 22 commands + ratatui TUI	`Cli`, `Commands`, `run_interactive_tui`	3
scheduler	Tokio-based cron automation	`ContribScheduler`	1
plugins	Trait-based plugin system	`AnalyzerPlugin`, `GeneratorPlugin`	1
templates	Contribution templates	`TemplateRegistry`	1
notifications	Slack/Discord/Telegram integrations	`Notifier`	1
sandbox	Code validation	`Sandbox`	1

Dependency Graph

                     ┌──────────────────┐
                     │   CLI / Web      │
                     │  (clap / axum)   │
                     └────────┬─────────┘
                              │
                   ┌──────────┴──────────┐
                   ▼                     ▼
            ┌─────────────┐      ┌──────────────┐
            │ Orchestrator│      │  Scheduler   │
            │  + Pipeline │      │   (tokio)    │
            └─────┬───────┘      └──────────────┘
                  │
        ┌─────────┼─────────┬──────────┐
        ▼         ▼         ▼          ▼
    ┌────────┐┌────────┐┌────────┐┌────────┐
    │Analysis││Generator││  PR    ││ Issues │
    │+Triage ││+Scorer ││Manager ││ Solver │
    └────┬───┘└────┬───┘└────┬───┘└────┬───┘
         │         │         │         │
         └─────────┼─────────┴─────────┘
                   │
        ┌──────────┴──────────┐
        ▼                     ▼
    ┌─────────┐         ┌──────────┐
    │   LLM   │         │  GitHub  │
    │ Routing │         │  Client  │
    └────┬────┘         └────┬─────┘
         │                   │
         └───────┬───────────┘
                 ▼
         ┌──────────────┐
         │    CORE      │
         │ (Models,     │
         │  Config,     │
         │  Middleware,  │
         │  Events)     │
         └──────────────┘

Dependency Flow: core ← github/llm ← analysis/generator ← orchestrator ← cli/web

Key Entry Points

CLI Entry Point

File: crates/contribai-rs/src/cli/mod.rs
Struct: Cli (clap derive)
22 Commands:
- run — Single full pipeline run
- hunt — Autonomous multi-round hunting
- patrol — Monitor open PRs
- target — Analyze specific repo
- analyze — Dry-run analysis
- solve — Solve issues in a repo
- stats — Summary statistics
- status — Show PR status table
- leaderboard — Merge rates by repo
- models — Available LLM models
- templates — Contribution templates
- profile — Named config profiles
- cleanup — Remove stale forks
- notify-test — Real HTTP to Slack/Discord/Telegram
- system-status — DB, rate limits, scheduler
- interactive — ratatui TUI browser (NEW in v5.2.0)
- web-server — Start web dashboard
- schedule — Start cron scheduler
- mcp-server — MCP stdio server
- init — Interactive setup wizard
- login — Auth status
- config-get/set/list — YAML config editor

Web Entry Point

File: crates/contribai-rs/src/web/mod.rs
Framework: Axum (tokio-based)
Key Routes:
- GET / — Dashboard HTML
- GET /api/stats — Overall statistics
- GET /api/repos — Analyzed repos list
- POST /api/run — Trigger pipeline (API key required)
- POST /api/run/target — Target specific repo (API key required)
- POST /api/webhooks/github — GitHub webhook (HMAC-SHA256)
- GET /api/health — Health check

MCP Server Entry Point

File: crates/contribai-rs/src/mcp/server.rs
Protocol: stdio JSON-RPC (Model Context Protocol)
Exposed Tools: 21 (GitHub read/write, safety, maintenance, PR management)

Critical Data Structures

Core Models (serde + struct)

Struct	Purpose	Key Fields
`Repository`	GitHub repo metadata	`owner`, `name`, `full_name`, `language`, `stars`, `forks`, `topics`
`Finding`	Detected issue	`finding_type`, `file_path`, `line`, `description`, `severity`, `context`
`Contribution`	Proposed fix	`finding`, `code_change`, `explanation`, `confidence_score`
`PRResult`	PR outcome	`pr_number`, `url`, `status`
`ContribAIConfig`	Application config	`github`, `llm`, `discovery`, `analysis`, `pipeline`, `web`

Database Schema (SQLite via rusqlite, 7 tables)

Table	Purpose
`analyzed_repos`	Track analyzed repositories
`submitted_prs`	All created PRs
`findings_cache`	Cached analysis results (72h TTL)
`run_log`	Pipeline execution history
`pr_outcomes`	PR merge/close outcomes for learning
`repo_preferences`	Learned repo patterns
`ci_monitor`	CI status tracking

Event Types (18 total)

RepositoryDiscovered | RepositoryAnalyzed | FindingDetected |
ContributionGenerated | PRCreated | PRMerged | PRClosed |
PRPatrolStarted | ReviewFound | CodeChangeGenerated |
ConfigLoaded | PipelineStarted | PipelineCompleted |
ErrorOccurred | RateLimitExceeded | IssueFound |
SchedulerStarted | WebhookReceived

Technology Stack

Category	Technologies
Language	Rust 2021 edition
Async Runtime	Tokio (multi-threaded)
Web	Axum, tower, hyper
HTTP Client	reqwest (async)
GitHub	reqwest + REST/GraphQL APIs
LLM	reqwest (Gemini, OpenAI, Anthropic, Ollama)
Serialization	serde, serde_json, serde_yaml
Database	rusqlite (sync, wrapped in tokio::task::spawn_blocking)
CLI	clap (derive macros)
Logging	tracing, tracing-subscriber
Code Parsing	tree-sitter (8 language grammars)
Crypto	hmac, sha2, hex (webhook verification)
Testing	cargo test (built-in), 335 tests
Linting	clippy
Formatting	rustfmt

Rust-Only Features (Not in Python)

Feature	Module	Description
Tree-sitter AST	`analysis/ast_intel.rs`	Parse 8 languages (Rust, Python, JS, TS, Go, Java, C, C++)
PageRank file ranking	`analysis/repo_map.rs`	Rank file importance via import graph analysis
12-signal triage	`analysis/triage.rs`	Score issues by recency, complexity, maintainer activity, etc.
3-tier context compression	`analysis/compressor.rs`	Language-aware signature extraction for 5 languages
Language rules	`analysis/language_rules.rs`	Per-language analysis rules and patterns
Leaderboard	`core/leaderboard.rs`	Contribution tracking and ranking

Common Code Patterns

Pattern 1: Async with Tokio

pub async fn process_repo(&self, repo: &Repository) -> Result<PipelineResult> {
    let findings = self.analyzer.analyze(repo).await?;
    let contributions = self.generator.generate_fixes(&findings).await?;
    let prs = self.pr_manager.create_prs(repo, &contributions).await?;
    Ok(PipelineResult { repo: repo.clone(), prs })
}

Pattern 2: Serde Models

#[derive(Debug, Clone, Serialize, Deserialize)]
pub struct Finding {
    pub finding_type: String,
    pub file_path: String,
    pub line: usize,
    pub description: String,
    pub severity: Severity,
}

Pattern 3: Trait-Based Providers

pub trait LlmProvider: Send + Sync {
    async fn complete(&self, prompt: &str) -> Result<String>;
    fn name(&self) -> &str;
}

Pattern 4: Middleware Chain

pub struct MiddlewareChain {
    middlewares: Vec<Box<dyn Middleware>>,
}
// RateLimit → Validation → Retry → DCO → QualityGate

Pattern 5: Event Bus

event_bus.emit(Event::PRCreated {
    repo: repo.full_name.clone(),
    pr_number,
    url: pr_url.clone(),
    timestamp: Utc::now(),
});

Testing Structure

Tests are co-located in each source file using #[cfg(test)] mod tests:

#[cfg(test)]
mod tests {
    use super::*;

    #[test]
    fn test_feature() {
        // Arrange, Act, Assert
    }

    #[tokio::test]
    async fn test_async_feature() {
        // Async test with tokio runtime
    }
}

Test Coverage: 335 tests across 63 source files Test Command: cargo test (all tests), cargo test <module> (specific)

Configuration

Source: crates/contribai-rs/src/core/config.rs
Format: YAML (config.yaml) + environment variables (CONTRIBAI_*)
Load Order: CLI flags → Env vars → YAML → Defaults
Key Structs: ContribAIConfig, GitHubConfig, LlmConfig, DiscoveryConfig, WebConfig

Document Metadata

Created: 2026-03-28
Last Updated: 2026-04-01
Version: 5.2.0 (Interactive TUI, real notify-test, full CLI parity)

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

ContribAI Codebase Summary

Quick Navigation

Module Responsibilities

Dependency Graph

Key Entry Points

CLI Entry Point

Web Entry Point

MCP Server Entry Point

Critical Data Structures

Core Models (serde + struct)

Database Schema (SQLite via rusqlite, 7 tables)

Event Types (18 total)

Technology Stack

Rust-Only Features (Not in Python)

Common Code Patterns

Pattern 1: Async with Tokio

Pattern 2: Serde Models

Pattern 3: Trait-Based Providers

Pattern 4: Middleware Chain

Pattern 5: Event Bus

Testing Structure

Configuration

Document Metadata

FilesExpand file tree

codebase-summary.md

Latest commit

History

codebase-summary.md

File metadata and controls

ContribAI Codebase Summary

Quick Navigation

Module Responsibilities

Dependency Graph

Key Entry Points

CLI Entry Point

Web Entry Point

MCP Server Entry Point

Critical Data Structures

Core Models (serde + struct)

Database Schema (SQLite via rusqlite, 7 tables)

Event Types (18 total)

Technology Stack

Rust-Only Features (Not in Python)

Common Code Patterns

Pattern 1: Async with Tokio

Pattern 2: Serde Models

Pattern 3: Trait-Based Providers

Pattern 4: Middleware Chain

Pattern 5: Event Bus

Testing Structure

Configuration

Document Metadata