Jev Scout — Development agent for Claude Code
Fast read-only scout for Jev-routed subtasks.
How to install Jev Scout
Installs to ~/.claude/agents/dansya-arsana-jev-harness-jev-scout.md
mkdir -p ~/.claude/agents && curl -fsSL https://raw.githubusercontent.com/dansya-arsana/jev-harness/HEAD/agents/jev-scout.md -o ~/.claude/agents/dansya-arsana-jev-harness-jev-scout.md Restart Claude Code, or start a new session, for it to be picked up.
What Jev Scout does
name: jev-scout description: Fast read-only scout for Jev-routed subtasks. Finds files, reads code, answers "where is X / what does Y import". Never edits. model: claude-opus-5-5 effort: low tools: Read, Grep, Glob, Bash
You are a read-only scout. Never create, edit, or delete files, and never run commands that change state (no installs, writes, git commits, or network calls that modify anything).
Answer the question in the prompt and stop. Report back briefly:
- the answer, with `path
Alternatives in Development
- Swarm Worker — Executes subtasks in a swarm - fast, focused, cost-effective 508 ★
- Book Genesis Auditor — Read-only structural auditor for the book-genesis skill 139 ★
- Ds Flash — DSH (DeepSeek Harness) worker on the flash tier - fast and cheap 116 ★
Full documentation available on GitHub
View Source RepositoryRelated Agents
Jev Builder
Implements clear, well-specified edits in one or two files for Jev-routed subtasks.
Jev Engineer
Implements coordinated multi-file changes with a known approach for Jev-routed subtasks (feature slices, API p
Scout Diagnose
Read-only bug investigation. Given a reproduced failure and a hypothesis, finds where it fails, why, and what
Eval Case Auditor
Read-only auditor for triage eval cases. Use after adding several cases or before promoting a baseline. Report
Haiku Scout
Haiku 4.5 read-only scout. Locate definitions and call sites, map a directory, condense a long log. Never edit
Next Issue Finder
Scout for undirected /orc runs — surveys the GitHub issue board, picks the next issue with orc's four-signal r
Related Skills
Uc Touchpoints
Print the advisory touchpoint floor for one opaque use-case id (UC-NNNN), grouped by SDLC stage — refine, solu
Multi Agent Orchestrator
Multi-model orchestrator: runs Claude Code, Codex, Gemini, DeepSeek, GPT, Qwen and more as interchangeable wor
Next Stage
Advance one gated issue to its next stage — spec refinement (2), parallel impl+QA (3+4+5), or the fast track —