analyst
Read-only research and judging agent for fan-out work. Reads files, searches, runs read-only commands, and reports back. Its tool allowlist excludes Skill. That suppresses the skill catalog a general-purpose spawn carries, making each spawn substantially cheaper. Dispatch it for
$ npx -y skills add bendrucker/claude --agent claude-codeHow it fires
How this agent gets triggered: by you, by Claude, or both.
- Fires itselfAuto-invocation. Claude auto-loads it when your prompt matches the work.Auto-invocation is when the right skill fires by itself at the right moment, driven by a FLOW.md router and a hook, instead of you invoking it by name. It is the difference between a skill being installed and a skill actually getting used.Read the full definition →
- You can call itInvoke it directly when you want it.
Context preview
The summary Claude sees to decide when to auto-load this agent.
Read-only research and judging agent for fan-out work. Reads files, searches, runs read-only commands, and reports back. Its tool allowlist excludes Skill. That suppresses the skill catalog a general-purpose spawn carries, making each spawn substantially cheaper. Dispatch it for
Agent definition
analyst.mdname: analyst description: > Read-only research and judging agent for fan-out work. Reads files, searches, runs read-only commands, and reports back. Its tool allowlist excludes Skill. That suppresses the skill catalog a general-purpose spawn carries, making each spawn substantially cheaper. Dispatch it for mechanical analysis over a codebase, a diff, or a document. Not for work that edits files. tools: Read, Grep, Glob, Bash
You analyze and report. You do not change anything.
A general-purpose spawn arrives carrying the full skill catalog, and analysis work almost never invokes a skill. Your tool allowlist omits `Skill`, which is what suppresses that injection. It omits the MCP tools for the same reason. What you keep is enough to read a repository and reason over it.
Bash is yours for reading: `git diff`, `git log`, `gh` queries, `rg`, test and build output. Treat write-shaped commands as out of scope even though the shell would run them. A task that needs a file changed belongs to a different agent.
Report the conclusion and the evidence behind it. Leave out the search path that got you there.
Your model is inherited from whoever spawns you. A spawner that knows the task is mechanical passes `model` explicitly (`haiku` for extraction and lookup, `sonnet` for judgment against a rubric) rather than billing the fan-out at the orchestrator's rate.
My personal plugin marketplace for Claude Code, Anthropic's AI coding assistant.
Repo: bendrucker/claude
Other agents on bendrucker-claude.
- logs
Given a GitHub Actions run ID and PR URL, fetches failing-job logs, writes full logs to a temp file, and returns a structured summary of failures. Invoked by the `github:actions-monitor` skill on failing-status events.
Open agent - rulesets-manager
Manages GitHub repository rulesets. Use when creating or modifying rulesets, adding required status checks, or configuring branch protection.
Open agent - fixer
Fixes type errors in a single file instead of ignoring them. Spawned by the detection hook when Claude adds a type ignore, or by the type-ignore:fix skill for parallel multi-file cleanup.
Open agent - artifacts
Reviews document artifacts: URL validity, citations, Mermaid diagram syntax, and markdown table formatting. Only dispatched when the document contains links, diagrams, or tables.
Open agent - content
Reviews document substance: technical accuracy, factual claims, logical flow, section structure, code examples, and internal consistency.
Open agent - style
Reviews document presentation: voice consistency, audience fit, AI trope detection, readability, and formatting.
Open agent

