rdf4j

Run Maven tests in this repo with a consistent workflow (module test-artifact cleanup, root -Pquick install to refresh .m2_repo, then module verify or a single test class/method). Use when asked to run tests/verify in the rdf4j multi-module build or when the user says mvnf.

Use QueryPlanSnapshotCli to capture and compare RDF4J query plans, then assess likely performance improvements/regressions from execution verification and semantic plan diffs. Trigger when users ask about optimizer impact, query-plan drift, join algorithm changes, or query performance regressions in testsuites/benchmark.

Diagnose Java performance issues by inspecting HotSpot bytecode, tiered compilation state, inlining decisions, C2 assembly, compilation limits, and object layout. Produces reproducible artifacts (jit.xml, directives, JFR, disassembly) and links them to actionable code changes.

Run Maven tests in this repo with a consistent workflow (module test-artifact cleanup, root -Pquick install to refresh .m2_repo, then module verify or a single test class/method). Use when asked to run tests/verify in the rdf4j multi-module build or when the user says mvnf.

Use QueryPlanSnapshotCli to capture and compare RDF4J query plans, then assess likely performance improvements/regressions from execution verification and semantic plan diffs. Trigger when users ask about optimizer impact, query-plan drift, join algorithm changes, or query performance regressions in testsuites/benchmark.

Run a repeatable RDF4J performance loop against one JMH benchmark in Docker with Linux Java 26 and JFR CPU-time profiling. Use when working in this repo on benchmark-guided performance changes, hotspot triage, JFR reading, CPU bottleneck analysis, or repeated baseline, fix, and rerun loops. Trigger on requests mentioning benchmark, profiling, JFR, hotspot, perf loop, CPU bottleneck, or Docker benchmark runs in RDF4J.

Parse JMH result text by finding the first header line that starts with Benchmark and contains Mode and Score, build a structured table for all columns/rows, compare overlapping benchmarks across 2+ files, compute Diff Score and Diff %, filter by deviation or regression thresholds, analyze regressions over time from filename/mtime timestamps, and export sortable reports to txt/md/csv/xlsx/html. Use for benchmark run comparisons, regression triage, and directory-wide historical analysis.

Use when writing, reviewing, or reshaping HotSpot Java where algorithmic complexity, data-structure choice, throughput, latency, allocation rate, zero-copy, lazy evaluation, non-materialization, runtime specialization, query-engine code generation, Janino, primitive collections, performance libraries, intrinsics, SuperWord auto-vectorization, or C2 assembly matter. Also use for advanced algorithmic problem solving in Java, including dynamic programming, graph/range techniques, cache-aware code shape, and choosing between interpreted, vectorized, and compiled execution paths. Bias toward asymptotic wins first, then the right execution model, then specialized hot-path code, then benchmark and JIT evidence.

Retrieve GitHub issues, pull requests, and milestones with read-only, whitelisted `gh` commands only. Use when you need complete issue or PR context, need to resolve a PR from commit ID/PR ID/issue ID, fetch milestone metadata, or list all issues in a milestone (labels, status, assignees, and related fields).

Debug Maven Surefire unit tests by running them in JDWP "wait for debugger" mode (`-Dmaven.surefire.debug`) and attaching to the forked test JVM using **jdb** (preferred for CLI/agent debugging), IntelliJ, or VS Code. Use when asked to debug/step through a failing JUnit test, attach a debugger to a Maven test run, or run `mvn test -Dtest=Class[#method]` suspended on a port (including multi-module `-pl` runs). The JVM will block at startup until a debugger attaches; the agent should attach with `jdb -attach <host>:<port>` and drive the session from the terminal.