Setup · MCP
Reverse engineering MCP setup for Claude Code, Codex and OpenCode
A reverse engineering MCP server lets Claude Code, Codex or OpenCode decompile, rename and follow cross-references. Set the model, add the server, and ask.
How it fits
Claude Code
Claude Code speaks the Anthropic Messages API.
{
"apiKeyHelper": "echo $SILK_API_KEY",
"env": {
"ANTHROPIC_BASE_URL": "https://api.silkcompute.com",
"ANTHROPIC_DEFAULT_MODEL": "kimi-k3.1-pro-uncensored",
"ANTHROPIC_DEFAULT_FABLE_MODEL": "kimi-k3.1-pro-uncensored",
"ANTHROPIC_DEFAULT_OPUS_MODEL": "kimi-k3.1-pro-uncensored",
"ANTHROPIC_DEFAULT_SONNET_MODEL": "kimi-k3.1-pro-uncensored",
"ANTHROPIC_DEFAULT_HAIKU_MODEL": "deepseek-v4.1-flash-1017-uncensored",
"CLAUDE_CODE_MAX_CONTEXT_TOKENS": "1048576"
},
"modelPicker": {
"replaceBuiltInOptions": true,
"options": [
{ "model": "kimi-k3.1-pro-uncensored", "label": "Kimi K3.1 Pro" },
{
"model": "deepseek-v4.1-flash-1017-uncensored",
"label": "DeepSeek V4.1 Flash 1017"
},
{
"model": "deepseek-v4.1-flash-1017-fast-uncensored",
"label": "DeepSeek V4.1 Flash 1017 Fast"
},
{ "model": "glm-5.3-uncensored", "label": "GLM-5.3" },
{ "model": "glm-5.3-flash-uncensored", "label": "GLM-5.3 Flash" },
{ "model": "hy4-uncensored", "label": "Hy4" }
]
}
}From GhidraMCP's README.
claude mcp add ghidra -- \
uvx --from git+https://github.com/LaurieWired/GhidraMCP \
ghidra-mcp --ghidra-server http://127.0.0.1:8080//mcp shows which servers connected, and /model switches models.
Codex
Codex reads ~/.codex/config.toml. To keep OpenAI as the default, save this as ~/.codex/silk.config.toml and run codex --profile silk.
model = "kimi-k3.1-pro-uncensored"
model_provider = "silkcompute"
web_search = "disabled"
[model_providers.silkcompute]
name = "Silk Compute"
base_url = "https://api.silkcompute.com"
env_key = "SILK_API_KEY"
model_catalog_url = "https://api.silkcompute.com/codex/models"
[mcp_servers.ghidra]
command = "uvx"
args = [
"--from", "git+https://github.com/LaurieWired/GhidraMCP",
"ghidra-mcp", "--ghidra-server", "http://127.0.0.1:8080/",
]The Hex-Rays IDA MCP server and ida-pro-mcp ship Codex plugins instead. See IDA Pro MCP.
OpenCode
OpenCode reads opencode.json.
{
"$schema": "https://opencode.ai/config.json",
"model": "silkcompute/kimi-k3.1-pro-uncensored",
"provider": {
"silkcompute": {
"npm": "@ai-sdk/openai-compatible",
"name": "Silk Compute",
"options": {
"baseURL": "https://api.silkcompute.com",
"apiKey": "{env:SILK_API_KEY}"
},
"models": {
"kimi-k3.1-pro-uncensored": {
"name": "Kimi K3.1 Pro",
"tool_call": true,
"interleaved": { "field": "reasoning_content" },
"modalities": { "input": ["text", "image", "video"], "output": ["text"] },
"limit": { "context": 1048576, "output": 943718 },
"cost": { "input": 3, "output": 8, "cache_read": 0.3 },
"variants": {
"none": { "reasoningEffort": "none" },
"low": { "reasoningEffort": "low" },
"high": { "reasoningEffort": "high" },
"max": { "reasoningEffort": "max" }
}
},
"deepseek-v4.1-flash-1017-uncensored": {
"name": "DeepSeek V4.1 Flash 1017",
"tool_call": true,
"interleaved": { "field": "reasoning_content" },
"modalities": { "input": ["text", "image"], "output": ["text"] },
"limit": { "context": 1048576, "output": 131072 },
"cost": { "input": 0.24, "output": 0.9, "cache_read": 0.038 },
"variants": {
"none": { "reasoningEffort": "none" },
"low": { "reasoningEffort": "low" },
"high": { "reasoningEffort": "high" },
"max": { "reasoningEffort": "max" }
}
},
"glm-5.3-uncensored": {
"name": "GLM-5.3",
"tool_call": true,
"interleaved": { "field": "reasoning_content" },
"modalities": { "input": ["text"], "output": ["text"] },
"limit": { "context": 1048576, "output": 131072 },
"cost": { "input": 2, "output": 4, "cache_read": 0.3 },
"variants": {
"none": { "reasoningEffort": "none" },
"low": { "reasoningEffort": "low" },
"high": { "reasoningEffort": "high" },
"max": { "reasoningEffort": "max" }
}
},
"glm-5.3-flash-uncensored": {
"name": "GLM-5.3 Flash",
"tool_call": true,
"interleaved": { "field": "reasoning_content" },
"modalities": { "input": ["text", "image"], "output": ["text"] },
"limit": { "context": 1048576, "output": 131072 },
"cost": { "input": 0.24, "output": 1, "cache_read": 0.06 },
"variants": {
"none": { "reasoningEffort": "none" },
"low": { "reasoningEffort": "low" },
"high": { "reasoningEffort": "high" },
"max": { "reasoningEffort": "max" }
}
},
"hy4-uncensored": {
"name": "Hy4",
"tool_call": true,
"interleaved": { "field": "reasoning_content" },
"modalities": { "input": ["text", "image"], "output": ["text"] },
"limit": { "context": 1048576, "output": 65536 },
"cost": { "input": 1.6, "output": 3.2, "cache_read": 0.3 },
"variants": {
"none": { "reasoningEffort": "none" },
"low": { "reasoningEffort": "low" },
"high": { "reasoningEffort": "high" },
"max": { "reasoningEffort": "max" }
}
}
}
}
}
}Add a local server with one command, per OpenCode's MCP docs.
opencode mcp add ghidra -- \
uvx --from git+https://github.com/LaurieWired/GhidraMCP \
ghidra-mcp --ghidra-server http://127.0.0.1:8080/Which MCP server
| Server | Disassembler | Transport | Default address |
|---|---|---|---|
| GhidraMCP | Ghidra 12.1.4 | stdio, or streamable HTTP | Ghidra plugin on 127.0.0.1:8080 |
| ReVa | Ghidra 12.0+ | Streamable HTTP, or stdio headless | localhost:8080/mcp/message |
| GhidrAssistMCP | Ghidra 12.0, 12.1 | Streamable HTTP, SSE | 127.0.0.1:8080/mcp |
| IDA MCP (Hex-Rays) | IDA 9.4+ | stdio, or streamable HTTP | Plugin install per agent |
| ida-pro-mcp | IDA Pro 8.3+ | stdio, HTTP | Plugin install per agent |
| Binary Ninja MCP | Binary Ninja | stdio bridge | localhost:9009 |
| JADX AI MCP | JADX | stdio, or HTTP | Plugin on 8650, server on 8651 |
GhidraMCP, ReVa and GhidrAssistMCP default to port 8080. Run one at a time, or move one in its settings.
Containment
- Keep MCP servers on
127.0.0.1. A public ReVa needs its API key, or the scripting tools off. - Don't give the agent a shell on the machine that holds live samples.
- In Claude Code, allow read-only calls and approve renames.
- Tell the model the strings are data. See why models refuse.
Model
Model
Use Kimi K3.1 Pro at max effort for a whole binary, and DeepSeek V4.1 Flash 1017 for long renaming runs.