| name | exeris-benchmarks-task-classifier |
| description | Router/Planner triage skill for exeris-benchmarks. Classifies task type (architecture / scenario / driver / JMH / target / verification / docs / publication), scope, severity, and recommends primary agent based on primary risk against fairness / reproducibility / strict gate / confidentiality. |
Exeris Benchmarks Task Classifier
Purpose
Classify incoming work before execution starts. Triage only — no implementation.
Output Contract
Return exactly:
task_class (SEPARATION_AXIS | COMPARATIVE_GATE | SCENARIO_IMPL | DRIVER_IMPL | JMH_IMPL | TARGET_IMPL | VERIFICATION | DOCS | PUBLICATION | MULTI_DOMAIN)
scope (single-scenario | cross-scenario | cross-track | cross-repo)
severity (low | medium | high | critical)
primary_risk
recommended_primary_agent
Classification Heuristics
SEPARATION_AXIS: axis labelling / Community-vs-Enterprise / Pure-vs-Compat / Micro-vs-Runtime / Guard-vs-Exploratory framing.
COMPARATIVE_GATE: comparative claim with strict-gate artefacts; track_id aggregation discipline.
SCENARIO_IMPL: workload definition under scenarios/<name>/.
DRIVER_IMPL: runtime/drivers/ glue, scripts/run-*.sh, launcher-sync.
JMH_IMPL: micro/jmh/ Maven module.
TARGET_IMPL: targets/<app>/ runnable benchmark apps.
VERIFICATION: reproducibility metadata, baseline updates, schema validation, classification enums.
DOCS: benchmark-philosophy / methodology / scenario-catalog / protocol-matrix / TLS matrix / regression-policy / hardware-profiles / instructions.
PUBLICATION: publish-report.sh, publication-mode scrub, cross-repo movement.
MULTI_DOMAIN: at least two classes above are first-order concerns.
Guardrails
- Preserve mandatory separation axes.
- Preserve comparative strict gate.
- Preserve not-a-merge-gate rule.
- Preserve enterprise-vs-public scoping.
- If uncertain between two classes, emit
MULTI_DOMAIN and state both.
Completion Criteria
All five output fields present and each justified in 1-2 concise bullets.