Skill: Gradle MCP Progress Reporting Expert

This skill provides expert guidance on implementing and managing progress reporting within the Gradle MCP project, including patterns for background builds and long-running operations.

Core Reporting Guidelines

• Always Report Progress: ALWAYS send progress commands/notifications (

  ProgressReporter

  ) when implementing long-running operations or tool handlers.
• Initial Feedback: Always report an initial "Starting..." or "Preparing..." progress status (0.0) before invoking a long-running operation to prevent the user from perceiving the tool as "stuck".
• Reporting Frequency: Do NOT artificially limit reports (e.g.,

  if count % 100

  ) in trackers or producers; throttled delivery is handled authoritatively at the top level.
• Accuracy Policy: Progress reporting does not have to be perfect—some slight or temporary inaccuracies are acceptable. Focus on overall user experience.

Advanced Patterns

• Parallel Strategy: When designing progress for parallel operations, prioritize stable activity messaging and independent phase ranges over jittery percentages.
• Merging Progress: Base merging progress across multiple search providers on the total document count across all providers rather than the number of providers themselves.
• Internal Gradle Progress: For long-running Gradle tasks in init scripts, explicitly report progress during slow internal operations (e.g., dependency resolution) using the format:

  [gradle-mcp] [PROGRESS] [CATEGORY]: CURRENT/TOTAL: MESSAGE

  .
• Filtering Differentiation: When filtering dependencies in init scripts, distinguish between "skipped by filter" and "up-to-date" to prevent false reporting.
• Job Management: When managing background collection jobs, always use

  job.cancelAndJoin()

  to ensure clean termination.

Thread Safety & Consistency

• Eventual Consistency: Progress trackers handle updates from multiple Gradle listener threads. Eventual consistency is acceptable; avoid heavy synchronization.
• State Consolidation: Prefer consolidating sub-task progress into a single state object within the tracker.