.coderabbit.yml

# yaml-language-server: $schema=https://coderabbit.ai/integrations/schema.v2.json
# CodeRabbit Configuration for Stringy
# Schema: https://coderabbit.ai/integrations/schema.v2.json
# Documentation: https://docs.coderabbit.ai/getting-started/yaml-configuration

# =============================================================================
# GENERAL SETTINGS
# =============================================================================

language: "en-US"

tone_instructions: "Be direct and technical. Focus on correctness, performance, and idiomatic Rust patterns. Stringy is a binary analysis tool - prioritize security considerations and false positive reduction."

early_access: true
enable_free_tier: true
inheritance: false

# =============================================================================
# REVIEWS CONFIGURATION
# =============================================================================
reviews:
    profile: "assertive"
    request_changes_workflow: false
    high_level_summary: true
    high_level_summary_instructions: "Focus on architectural impact, API changes, binary format handling, and semantic classification accuracy."
    high_level_summary_in_walkthrough: true
    review_status: true
    commit_status: true
    fail_commit_status: false
    collapse_walkthrough: true
    changed_files_summary: true
    sequence_diagrams: true
    estimate_code_review_effort: true
    assess_linked_issues: true
    related_issues: true
    related_prs: true
    suggested_labels: true
    auto_apply_labels: true
    suggested_reviewers: true
    auto_assign_reviewers: true
    poem: true
    abort_on_close: true
    disable_cache: false

    # ---------------------------------------------------------------------------
    # Path Filters
    # ---------------------------------------------------------------------------
    path_filters:
        - "!**/Cargo.lock"
        - "!**/target/**"
        - "!**/.git/**"
        - "!**/tests/fixtures/**"
        - "!**/tests/snapshots/**"

    # ---------------------------------------------------------------------------
    # Path Instructions - Module-specific review guidelines
    # ---------------------------------------------------------------------------
    path_instructions:
        # -------------------------------------------------------------------------
        # Core Library
        # -------------------------------------------------------------------------
        - path: "src/lib.rs"
          instructions: |
              This is the library entry point with module declarations and re-exports. Review for:
              - #![forbid(unsafe_code)] and #![deny(warnings)] are present
              - Public re-exports are ergonomic (import from stringy::types, not deeply nested)
              - Module organization matches the data flow pipeline
              - No unnecessary public exports

        - path: "src/main.rs"
          instructions: |
              This is the CLI entry point using clap derive macros. Review for:
              - Correct use of clap attributes (#[arg], #[command])
              - Help text is clear and actionable
              - Error handling uses Box<dyn std::error::Error> or StringyError
              - Pipeline integration: format detection -> parsing -> extraction -> classification -> output

        - path: "src/types.rs"
          instructions: |
              This file defines core data structures. Review for:
              - All public structs use #[non_exhaustive] with explicit constructors
              - Tag enum covers all semantic classification types
              - StringyError uses thiserror with descriptive messages and context
              - Encoding, BinaryFormat, SectionType, StringSource enums are complete
              - FoundString and ContainerInfo have all necessary metadata fields
              - All structs derive appropriate traits (Debug, Clone, Serialize, Deserialize)

        # -------------------------------------------------------------------------
        # Container Module - Binary Format Parsing
        # -------------------------------------------------------------------------
        - path: "src/container/mod.rs"
          instructions: |
              This file defines the ContainerParser trait and format detection. Review for:
              - ContainerParser trait has detect() and parse() methods
              - detect_format() correctly identifies ELF, PE, Mach-O via magic bytes
              - create_parser() returns appropriate parser for each format
              - Error handling uses StringyError variants

        - path: "src/container/elf.rs"
          instructions: |
              This file implements ELF binary parsing via goblin. Review for:
              - Section weight system (1.0-10.0) follows established patterns:
                - .rodata: 10.0, .comment/.note: 9.0, .dynstr/.strtab: 8.0
                - .data.rel.ro: 7.0, .data: 5.0
              - Import/export extraction from symbol tables
              - SectionType classification is accurate
              - Error context includes section names and offsets

        - path: "src/container/pe.rs"
          instructions: |
              This file implements PE binary parsing via goblin. Review for:
              - Section weight system follows established patterns:
                - .rdata: 10.0, .rsrc: 9.0, .data (by permissions): 5.0-7.0, .text: 3.0
              - Import/export extraction from PE tables
              - Resource section handling coordinates with pe_resources.rs
              - Windows-specific characteristics are handled

        - path: "src/container/macho.rs"
          instructions: |
              This file implements Mach-O binary parsing via goblin. Review for:
              - Segment/section weight system follows established patterns:
                - __TEXT,__cstring: 10.0, __TEXT,__const: 9.0
                - __DATA_CONST: 7.0, __DATA,__data: 5.0
              - Load command processing
              - Universal binary (fat) handling if applicable
              - Import/export extraction from symbol tables

        # -------------------------------------------------------------------------
        # Extraction Module - String Extraction
        # -------------------------------------------------------------------------
        - path: "src/extraction/mod.rs"
          instructions: |
              This file defines the StringExtractor trait and BasicExtractor. Review for:
              - StringExtractor trait signature: extract(&self, data, info) -> Vec<FoundString>
              - BasicExtractor combines ASCII, UTF-16, and filter configs
              - Per-section extraction respects section weights
              - Proper coordination between extraction components

        - path: "src/extraction/ascii.rs"
          instructions: |
              This file handles ASCII/UTF-8 string extraction. Review for:
              - AsciiExtractionConfig has sensible defaults (min_length: 4, max_length: 1024)
              - UTF-8 validation is correct
              - Confidence scoring logic is sound
              - Performance for large binaries

        - path: "src/extraction/utf16.rs"
          instructions: |
              This file handles UTF-16LE/BE string extraction. Review for:
              - Both little-endian and big-endian support
              - BOM detection and handling
              - Confidence scoring distinguishes real UTF-16 from garbage
              - Null-interleaved ASCII detection
              - Performance considerations

        - path: "src/extraction/dedup.rs"
          instructions: |
              This file handles string deduplication with occurrence tracking. Review for:
              - Grouping by (text, encoding) tuple
              - All occurrences are preserved with metadata
              - Tag merging uses HashSet union
              - Score formula: base + occurrence_bonus + cross_section_bonus + multi_source_bonus + confidence_boost
              - No data loss during deduplication

        - path: "src/extraction/pe_resources.rs"
          instructions: |
              This file extracts PE resources using pelite. Review for:
              - VERSIONINFO parsing and string extraction
              - STRINGTABLE resource handling
              - MANIFEST XML parsing
              - Resource metadata capture (type, language, size)
              - Error handling for malformed resources

        - path: "src/extraction/filters.rs"
          instructions: |
              This file implements noise filtering to reduce false positives. Review for:
              - Filter criteria are well-justified
              - Entropy-based filtering uses appropriate thresholds
              - No legitimate strings are filtered incorrectly
              - FilterConfig is configurable

        - path: "src/extraction/config.rs"
          instructions: |
              This file defines extraction configuration structs. Review for:
              - All config structs have sensible defaults
              - Builder pattern if used
              - Validation of config values
              - Documentation of config options

        # -------------------------------------------------------------------------
        # Classification Module - Semantic Tagging
        # -------------------------------------------------------------------------
        - path: "src/classification/mod.rs"
          instructions: |
              This file exports the classification module. Review for:
              - SemanticClassifier is publicly exported
              - Module organization is clean

        - path: "src/classification/semantic.rs"
          instructions: |
              This file implements semantic classification with regex patterns. Review for:
              - Regex patterns use lazy_static! for caching
              - URL pattern handles safe characters correctly
              - Domain validation includes TLD checking
              - IPv4/IPv6 uses regex pre-filter + std::net::IpAddr validation
              - POSIX paths: absolute paths starting with /
              - Windows paths: drive letter + backslash, no consecutive backslashes
              - UNC paths: \\server\share format
              - Registry paths: HKEY_* and HK* abbreviations
              - False positive reduction logic is sound
              - No catastrophic backtracking in regex patterns
              - classify() method returns Vec<Tag>

        # -------------------------------------------------------------------------
        # Output Module - Formatters
        # -------------------------------------------------------------------------
        - path: "src/output/mod.rs"
          instructions: |
              This file defines output formatters. Review for:
              - OutputFormatter trait if defined
              - JSON output is valid and complete
              - Human-readable output is clear
              - YARA-friendly output follows YARA syntax

        # -------------------------------------------------------------------------
        # Tests
        # -------------------------------------------------------------------------
        - path: "tests/**/*.rs"
          instructions: |
              Review test files for:
              - Integration tests use fixtures from tests/fixtures/
              - Snapshot testing uses insta with assert_debug_snapshot!
              - Test names follow test_<function>_<scenario> pattern
              - Both success and error cases are covered
              - Performance tests have reasonable timeouts
              - No flaky tests (avoid timing-dependent assertions)
              - Helper functions like make_found_string() are reused

        - path: "benches/**/*.rs"
          instructions: |
              Review benchmark files for:
              - Uses Criterion framework correctly
              - Benchmarks measure meaningful operations (parsing, extraction)
              - Input sizes are realistic (actual binary fixtures)
              - No I/O in hot paths being measured
              - Results are reproducible

        # -------------------------------------------------------------------------
        # Configuration & Documentation
        # -------------------------------------------------------------------------
        - path: "Cargo.toml"
          instructions: |
              Review Cargo.toml changes for:
              - Version bumps follow semver
              - Edition is 2024, MSRV is 1.85+
              - New dependencies are necessary and well-maintained
              - Core deps: goblin, pelite, clap, regex, lazy_static, serde, thiserror
              - Dev deps: criterion, insta, tempfile
              - No unnecessary feature flags

        - path: "*.md"
          instructions: |
              Review documentation for:
              - ASCII only - no emojis, em-dashes, or Unicode punctuation
              - Accuracy with current implementation
              - Clear examples that work
              - Proper markdown formatting
              - No broken links

        - path: "justfile"
          instructions: |
              Review justfile changes for:
              - Cross-platform compatibility (Windows PowerShell / Unix bash)
              - Commands follow existing patterns
              - No breaking changes to common commands (check, test, lint, build)

    # ---------------------------------------------------------------------------
    # Auto Review Settings
    # ---------------------------------------------------------------------------
    auto_review:
        enabled: true
        auto_incremental_review: true
        ignore_title_keywords:
            - "WIP"
            - "DO NOT REVIEW"
            - "[skip ci]"
            - "[skip review]"
        labels: []
        drafts: false
        base_branches: []
        ignore_usernames:
            - "dependabot[bot]"
            - "renovate[bot]"

    # ---------------------------------------------------------------------------
    # Finishing Touches
    # ---------------------------------------------------------------------------
    finishing_touches:
        docstrings:
            enabled: true
        unit_tests:
            enabled: true

    # ---------------------------------------------------------------------------
    # Pre-merge Checks
    # ---------------------------------------------------------------------------
    pre_merge_checks:
        docstrings:
            mode: "warning"
            threshold: 70

        title:
            mode: "warning"
            requirements: |
                PR titles must follow Conventional Commits format:
                <type>(<scope>): <description>

                Types: feat, fix, docs, style, refactor, perf, test, build, ci, chore
                Scopes: container, extraction, classification, output, types, cli, deps, release

                Examples:
                - feat(classification): add GUID pattern detection
                - fix(extraction): handle malformed PE resources
                - docs(readme): update installation instructions
                - refactor(container): extract common section weight logic

        description:
            mode: "warning"

        issue_assessment:
            mode: "warning"

        custom_checks:
            - mode: "warning"
              name: "No Unsafe Code"
              instructions: |
                  Verify that no unsafe code is introduced:
                  1. No `unsafe` blocks or functions
                  2. No `#[allow(unsafe_code)]` attributes
                  3. The crate uses `#![forbid(unsafe_code)]`

            - mode: "warning"
              name: "Error Handling"
              instructions: |
                  Verify that all error handling follows the project conventions:
                  1. Use StringyError variants from types.rs, never raw strings
                  2. Add context with descriptive messages including offsets and section names
                  3. Map external errors (goblin, pelite, io) to appropriate StringyError variants
                  4. Use thiserror #[error] and #[from] attributes correctly

            - mode: "warning"
              name: "ASCII Only"
              instructions: |
                  Verify that no Unicode punctuation is introduced unless explicitly required:
                  1. No emojis in code or documentation
                  2. No em-dashes - use regular hyphens
                  3. No smart quotes - use straight quotes
                  4. No other non-ASCII characters in strings or comments

            - mode: "warning"
              name: "File Size Limit"
              instructions: |
                  Verify that files stay under 500 lines:
                  1. New files should be under 500 lines
                  2. If a file exceeds 500 lines, consider splitting into modules
                  3. No blanket #[allow] attributes on modules or files

            - mode: "warning"
              name: "Section Weight Consistency"
              instructions: |
                  For changes to container parsers, verify section weights are consistent:
                  1. Read-only string sections: 9.0-10.0
                  2. String tables: 8.0
                  3. Read-only data: 7.0
                  4. Writable data: 5.0
                  5. Code sections: 3.0 or lower
                  6. Follow existing patterns in container/*.rs

    # ---------------------------------------------------------------------------
    # Labeling Instructions
    # ---------------------------------------------------------------------------
    labeling_instructions:
        - label: "bug"
          instructions: "Apply when the PR fixes something that is not working correctly. Look for fixes to parsing errors, incorrect classification, false positives/negatives, or crashes."
        - label: "enhancement"
          instructions: "Apply when the PR adds new features or improves existing functionality. This includes new binary format support, new semantic classifiers, new output formats, or performance improvements."
        - label: "documentation"
          instructions: "Apply when the PR primarily updates documentation files (*.md), code comments, or inline documentation."
        - label: "help wanted"
          instructions: "Apply when the PR is incomplete or the author explicitly requests help with implementation."
        - label: "question"
          instructions: "Apply when the PR raises questions about implementation approach or needs discussion before proceeding."
        - label: "duplicate"
          instructions: "Apply when the PR duplicates work from another open or merged PR."
        - label: "invalid"
          instructions: "Apply when the PR does not follow project conventions, targets the wrong branch, or is not appropriate for the project."

    # ---------------------------------------------------------------------------
    # Tools Configuration
    # ---------------------------------------------------------------------------
    tools:
        # Rust-specific tools
        clippy:
            enabled: true

        # General tools
        ast-grep:
            essential_rules: true
            rule_dirs: []
            util_dirs: []
            packages: []

        shellcheck:
            enabled: true

        markdownlint:
            enabled: true

        github-checks:
            enabled: true
            timeout_ms: 90000

        languagetool:
            enabled: true
            enabled_rules: []
            disabled_rules: []
            enabled_categories: []
            disabled_categories: []
            enabled_only: false

        gitleaks:
            enabled: true

        checkov:
            enabled: true

        semgrep:
            enabled: true

        actionlint:
            enabled: true

        yamllint:
            enabled: true

        # Disable tools not relevant to this project
        ruff:
            enabled: false
        eslint:
            enabled: false
        biome:
            enabled: false
        phpstan:
            enabled: false
        swiftlint:
            enabled: false
        rubocop:
            enabled: false
        detekt:
            enabled: false
        golangci-lint:
            enabled: false

# =============================================================================
# CHAT CONFIGURATION
# =============================================================================
chat:
    art: true
    auto_reply: true
    integrations:
        jira:
            usage: "disabled"
        linear:
            usage: "disabled"

# =============================================================================
# KNOWLEDGE BASE CONFIGURATION
# =============================================================================
knowledge_base:
    opt_out: false

    web_search:
        enabled: true

    code_guidelines:
        enabled: true
        filePatterns:
            - "AGENTS.md"
            - "CLAUDE.md"
            - ".github/copilot-instructions.md"
            - "README.md"
            - "codebase_analysis.md"

    learnings:
        scope: "local"

    issues:
        scope: "local"

    pull_requests:
        scope: "local"

    jira:
        usage: "disabled"
        project_keys: []

    linear:
        usage: "disabled"
        team_keys: []

    mcp:
        usage: "disabled"
        disabled_servers: []

# =============================================================================
# CODE GENERATION CONFIGURATION
# =============================================================================
code_generation:
    docstrings:
        language: "en-US"
        path_instructions:
            - path: "src/**/*.rs"
              instructions: |
                  Generate Rust documentation comments (///) for public items.
                  Follow these conventions:
                  - Start with a brief one-line summary
                  - Add # Examples section with working code when helpful
                  - Add # Errors section documenting error conditions
                  - Add # Panics section if the function can panic
                  - Use backticks for code references
                  - Reference related items with [links]
                  - ASCII only - no emojis or Unicode punctuation

    unit_tests:
        path_instructions:
            - path: "src/**/*.rs"
              instructions: |
                  Generate Rust unit tests following project conventions:
                  - Place tests in #[cfg(test)] mod tests { } blocks
                  - Use #[test] for all tests (project is synchronous)
                  - Use tempfile for filesystem fixtures
                  - Use insta for snapshot testing where appropriate
                  - Include both success and error cases
                  - Use assert_eq! with descriptive messages
                  - Name tests: test_<function>_<scenario>
                  - Binary fixtures go in tests/fixtures/

# =============================================================================
# ISSUE ENRICHMENT CONFIGURATION
# =============================================================================
issue_enrichment:
    auto_enrich:
        enabled: true

    planning:
        enabled: true
        auto_planning:
            enabled: false
            labels:
                - "enhancement"
                - "bug"

    labeling:
        auto_apply_labels: true
        labeling_instructions:
            - label: "bug"
              instructions: "Apply when the issue reports something that is not working correctly. Look for error messages, unexpected behavior, crashes, or incorrect classification results."
            - label: "enhancement"
              instructions: "Apply when the issue requests new features or improvements. This includes new binary format support, new semantic classifiers, new output formats, or performance improvements."
            - label: "documentation"
              instructions: "Apply when the issue is about missing, incorrect, or unclear documentation. This includes README updates, API documentation, examples, or inline code comments."
            - label: "good first issue"
              instructions: "Apply when the issue is well-scoped, has clear requirements, and does not require deep knowledge of the codebase. Good for newcomers to contribute."
            - label: "help wanted"
              instructions: "Apply when the issue needs community input, additional expertise, or the maintainers explicitly request assistance."
            - label: "question"
              instructions: "Apply when the issue is asking for clarification, guidance, or discussion rather than reporting a bug or requesting a feature."
            - label: "duplicate"
              instructions: "Apply when this issue duplicates an existing open or recently closed issue. Reference the original issue."
            - label: "invalid"
              instructions: "Apply when the issue does not provide enough information, is not related to this project, or cannot be reproduced."
            - label: "wontfix"
              instructions: "Apply when the issue describes behavior that is working as intended, is out of scope for the project, or conflicts with project goals."