Router Analyst — Research agent for Claude Code
Investigate router behaviour — a failing golden-set gate, a scorecard regression, a bucket that dropped, a capability the router stopped reaching, or a question about stage mix and latency.
How to install Router Analyst
Installs to ~/.claude/agents/andrecpramos-nextgen-interface-router-analyst.md
mkdir -p ~/.claude/agents && curl -fsSL https://raw.githubusercontent.com/andrecpramos/NextGen-Interface/HEAD/.claude/agents/router-analyst.md -o ~/.claude/agents/andrecpramos-nextgen-interface-router-analyst.md Restart Claude Code, or start a new session, for it to be picked up.
What Router Analyst does
name: router-analyst description: Investigate router behaviour — a failing golden-set gate, a scorecard regression, a bucket that dropped, a capability the router stopped reaching, or a question about stage mix and latency. Reads scorecards and runs the eval harness; recommends declaration changes and calibration, never hand-tuned thresholds. Use when accuracy moved or before shipping a registry change. tools: Read, Grep, Glob, Bash model: sonnet
You own the cascade's correctness. The r
Alternatives in Research
- Gemini Research Expert — Use this agent when the user needs to perform research tasks, gather information from external sources, or inv 36.6k ★
- Fable Worker Sonnet — Fable stage worker for tasks needing real reasoning — research synthesis, nontrivial code, analysis, document 837 ★
- Deep Research Writer — Writes the single report.md of a nav-deep-research run from the run's source notes, following the readable rep 232 ★
Full documentation available on GitHub
View Source RepositoryRelated Agents
Django API Architect Debugger
Investigate and fix Django REST Framework bugs, failing behaviour, and root causes using this repository's ser
Router Explorer
Investigate a bounded question about the codebase and return evidence. Read-only — locates code and reports wh
Mix Chain
Device configuration, mix chain decisions, plugin parameter reads and sets, spectrograph analysis. Read the re
Incident Triage
Investigate production incidents in Berserk — correlate errors, latency spikes, and log patterns across servic
Otel Trace
Investigate OpenTelemetry traces in Berserk — span analysis, latency debugging, trace correlation, service dep
Swarm Auditor
Fable 5.1 final auditor for research swarms — the last gate before a deliverable ships. Reads ONLY the distill
Related Skills
Prompt SLO
Prompt: Ship a latency SLO, as a ratio at a bucket boundary
Agent Reliability Harness
Evals, tracing and CI regression gates for tool-using agents: tool-selection, final-state, cost, latency and s
Eval Prompts
Run a golden-set prompt eval suite against the pinned baseline; fail on any regression beyond threshold.