Evaluator Fanout — Development agent for Claude Code
Spawn-capable per-surface evaluator — actuates the brain-owned child_fanout_plan through Claude's one supported teammate-to-subagent edge, one child per (bug_class × auth) cell.
How to install Evaluator Fanout
Installs to ~/.claude/agents/vmihalis-hacker-bob-evaluator-fanout.md
mkdir -p ~/.claude/agents && curl -fsSL https://raw.githubusercontent.com/vmihalis/hacker-bob/HEAD/.claude/agents/evaluator-fanout.md -o ~/.claude/agents/vmihalis-hacker-bob-evaluator-fanout.md Restart Claude Code, or start a new session, for it to be picked up.
What Evaluator Fanout does
name: evaluator-fanout description: Spawn-capable per-surface evaluator — actuates the brain-owned child_fanout_plan through Claude's one supported teammate-to-subagent edge, one child per (bug_class × auth) cell. Transition-blind: discovered cross-surface pivots ride discovered_pivots[] up to the orchestrator. tools: Bash, Read, Grep, Glob, Agent(evaluator-fanout-child), mcp__hacker-bob__bob_ingest_sarif, mcp__hacker-bob__bob_read_static_analysis_index, mcp__hacker-bob__bob_record_candidate
Alternatives in Development
- Part 13: Memory Bridge (Give Coding Agents Your Brain) — When you spawn Codex or Claude Code to build something, they start blind 232 ★
- Scope Guard (Shared Prompt Block For Tier 2 Agents) — This file is not a standalone agent 213 ★
- Franky — Implementation worker for one Superloopy criterion or one non-overlapping slice 109 ★
Full documentation available on GitHub
View Source RepositoryRelated Agents
Evaluator Fanout Child
Non-recursive evaluator-fanout leaf — tests one MCP-issued (bug_class × auth) cell, writes durable claim/cover
Disc Blind Reader
A blind reader of the stage-1 discovery review — reads the PR-FAQ and User Stories alone and commits to the co
Disc Reviewer Ambiguity
The ambiguity judge of the stage-1 discovery review — compares two blind readers' builds and reports every rea
Evaluator Spawn
Generic TaskGraph evaluator shell — executes Transition and Hypothesis nodes the orchestrator dispatched via b
Coder Teammate
Code implementation teammate. Works in Agent Teams mode, communicates directly with evaluator-teammate. Claims
Evaluator Teammate
Code evaluation teammate. Works in Agent Teams mode, receives Review Dispatch payloads, reviews scoped code ch
Related Skills
Orvima
Why should Claude be the only one with a browser? Orvima gives any MCP-capable AI (Claude, Cursor, Copilot, yo
Frontier Kb
Frontier intelligence KB: cross-harness research (OMP/Hermes/Codex/Claude/Cursor/Pi/…) + cutting-edge LLM/tool
Claude Teams Brain
Give your Claude Code Agent Teams a memory. Auto-injects role-specific context into every new teammate — your