rex

65 mods across 3 repositories, 2 stars between them.

debugger

01

rex/mcp-rancher

Agent Claude Code

Use PROACTIVELY for errors, stack traces, test failures, or unexpected behavior. Root-cause analysis + minimal fix.

1 6d ago A 29 tokens original MIT

implementer

02

rex/mcp-rancher

Agent Claude Code

Executes exactly ONE slice from TASKSTATE.md. Reads slice, writes code, runs tests, checks box. Never plans or decomposes. Invoke with slice ID.

1 6d ago A 36 tokens original MIT

planner

03

rex/mcp-rancher

Agent Claude Code

MUST BE USED for any feature, bug, or refactor touching >1 file or >30 LOC. Decomposes into verifiable slices in TASKSTATE.md. Planning only — does NOT implement.

1 6d ago A 43 tokens original MIT

rancher-tool-author

04

rex/mcp-rancher

Agent Claude Code

Use PROACTIVELY when authoring or extending MCP tools in tools/ /. Takes arguments domain (norman|steve|longhorn|fleet) and toolname. Loads ONLY the relevant domain's AGENTS.md and template.py into context. Drafts the tool implementation, tests, and TOOLCATALOG update.

1 6d ago A 76 tokens original MIT

research-agent

05

rex/mcp-rancher

Agent Claude Code

Use PROACTIVELY for any task requiring codebase search, doc fetch, or understanding how existing code works. Spawn multiple in parallel for independent questions. Returns terse structured synthesis — never raw dumps.

1 6d ago A 43 tokens original MIT

reviewer

06

rex/mcp-rancher

Agent Claude Code

Use PROACTIVELY after any implementer slice and before /ship. Reviews working-tree diff against CONVENTIONS.md. Read-only.

1 6d ago A 31 tokens original MIT

terraform-reviewer

07

rex/mcp-rancher

Agent Claude Code

Read-only Terraform reviewer. Use PROACTIVELY for any PR touching .tf, .tfvars, or modules/. Runs fmt/validate/lint/scan + plan summary. MUST NOT apply, destroy, or mutate state.

1 6d ago A 51 tokens original MIT

test-runner

08

rex/mcp-rancher

Agent Claude Code

Use PROACTIVELY after any code change. Runs project tests and reports failures with one-line root-cause guesses. Does not fix.

1 6d ago A 32 tokens original MIT

adr

09

rex/mcp-rancher

Command Claude Code

Draft a new Architectural Decision Record from the current discussion.

1 6d ago A 10 tokens original MIT

debug

10

rex/mcp-rancher

Command Claude Code

Root-cause a failing test or bug via the debugger subagent.

1 6d ago A 13 tokens original MIT

plan

12

rex/mcp-rancher

Command Claude Code

Decompose a feature into slices in TASKSTATE.md and specs/ /.

1 6d ago A 15 tokens original MIT

retrofit

13

rex/mcp-rancher

Command Claude Code

Retrofit an existing repo for agent collaboration (brownfield, 5 PRs).

1 6d ago C 17 tokens original MIT

review

14

rex/mcp-rancher

Command Claude Code

Run the reviewer subagent on the current working-tree diff.

1 6d ago A 11 tokens original MIT

scaffold

15

rex/mcp-rancher

Command Claude Code

Scaffold a fresh, agent-ready repo from scratch (greenfield bootstrap).

1 6d ago C 15 tokens original MIT

ship

16

rex/mcp-rancher

Command Claude Code

Pre-merge pipeline — test, review, update TASKSTATE, stage.

1 6d ago A 14 tokens original MIT

sync-skills

17

rex/mcp-rancher

Command Claude Code

Refresh skill-sourced .claude/ files from the installed agentic-skeleton templates.

1 6d ago B 18 tokens original MIT

terraform-plan

18

rex/mcp-rancher

Command Claude Code

Run the terraform-reviewer subagent on the current Terraform diff.

1 6d ago A 12 tokens original MIT

SessionStart

19

rex/mcp-rancher

Hook Claude Code

Runs when a session starts on startup, resume and clear, executing session-start.sh. From rex/mcp-rancher.

1 6d ago A tokens not measured copy · 48% MIT

UserPromptSubmit

20

rex/mcp-rancher

Hook Claude Code

Runs when you submit a prompt, before the agent sees it, executing serena-required.sh and inject-state.sh (2 commands). From rex/mcp-rancher.

1 6d ago A tokens not measured copy · 52% MIT

PreToolUse

21

rex/mcp-rancher

Hook Claude Code

Runs before the agent uses a tool for Bash, Bash, Read, Edit, MultiEdit, Write, Glob and Grep tool calls, executing bash-guard.sh and serena-gate.sh (2 commands). From rex/mcp-rancher.

1 6d ago A tokens not measured original MIT

PostToolUse

22

rex/mcp-rancher

Hook Claude Code

Runs after a tool call finishes for Edit, MultiEdit and Write tool calls, executing auto-lint.sh. From rex/mcp-rancher.

1 6d ago A tokens not measured copy · 52% MIT

PreCompact

23

rex/mcp-rancher

Hook Claude Code

Runs before the context is compacted, executing pre-compact.sh. From rex/mcp-rancher.

1 6d ago A tokens not measured copy · 89% MIT

Stop

24

rex/mcp-rancher

Hook Claude Code

Runs when the agent finishes a response, executing stop-gate.sh. From rex/mcp-rancher.

1 6d ago A tokens not measured copy · 47% MIT