{"passport":{"unfragile":{"@version":"1.0","version":"2026-05","artifact":{"id":"npm-codex-mcp-server","slug":"codex-mcp-server","name":"codex-mcp-server","type":"mcp","url":"https://github.com/tuannvm/codex-mcp-server#readme","page_url":"https://unfragile.ai/codex-mcp-server","categories":["mcp-servers"],"tags":["mcp","codex","openai","claude","ai","cli"],"pricing":{"model":"open_source","free":true,"starting_price":null},"status":"active","verified":false},"capabilities":[{"id":"npm-codex-mcp-server__cap_0","uri":"capability://tool.use.integration.mcp.protocol.wrapping.for.openai.codex.cli","name":"mcp protocol wrapping for openai codex cli","description":"Wraps the OpenAI Codex command-line interface as an MCP (Model Context Protocol) server, translating MCP tool calls into Codex CLI invocations and marshaling responses back through the MCP protocol. This enables Claude and other MCP-compatible clients to invoke Codex functionality through standardized tool-calling semantics without direct CLI knowledge.","intents":["I want Claude to use OpenAI Codex for code generation without managing CLI processes myself","I need to expose Codex capabilities as MCP tools in my Claude integration","I want to standardize how my AI agents interact with Codex through the MCP protocol"],"best_for":["developers building Claude-based coding assistants","teams integrating OpenAI Codex into MCP-compatible AI workflows","builders prototyping multi-model code generation pipelines"],"limitations":["Depends on OpenAI Codex CLI being installed and authenticated locally — no fallback if CLI unavailable","MCP server runs as a single process — no built-in clustering or load balancing for concurrent requests","Latency includes CLI subprocess overhead (process spawn + IPC) on top of Codex API latency","No request queuing or rate limiting — relies on upstream MCP client to manage concurrency"],"requires":["OpenAI Codex CLI installed and in PATH","OpenAI API key configured for Codex CLI authentication","Node.js 14+ (typical MCP server runtime requirement)","MCP-compatible client (Claude desktop, custom MCP client, etc.)"],"input_types":["text prompts","code context/snippets","structured tool parameters from MCP schema"],"output_types":["generated code strings","structured MCP tool results","error messages and status codes"],"categories":["tool-use-integration","code-generation-editing"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"npm-codex-mcp-server__cap_1","uri":"capability://automation.workflow.codex.cli.subprocess.orchestration","name":"codex cli subprocess orchestration","description":"Manages spawning, communication with, and lifecycle of OpenAI Codex CLI subprocesses. Handles stdin/stdout marshaling, error capture, and process cleanup to reliably invoke Codex operations from within the Node.js MCP server process without blocking or resource leaks.","intents":["I need to invoke Codex CLI commands from Node.js without blocking the event loop","I want reliable subprocess communication with proper error handling and cleanup","I need to capture Codex output and errors for structured response formatting"],"best_for":["Node.js developers wrapping CLI tools as MCP servers","teams building subprocess-based integrations with legacy command-line tools"],"limitations":["Subprocess spawning adds 50-200ms overhead per invocation depending on system load","No built-in timeout handling — long-running Codex operations can block MCP responses","Requires shell/OS-level process management — behavior may differ across Linux, macOS, Windows","No streaming support — entire Codex output must be buffered before returning to MCP client"],"requires":["Node.js child_process module (built-in)","OpenAI Codex CLI binary accessible in system PATH","Sufficient system resources for subprocess spawning"],"input_types":["CLI arguments and flags","stdin data (code context, prompts)"],"output_types":["stdout text (generated code)","stderr text (errors, warnings)","exit codes"],"categories":["automation-workflow","tool-use-integration"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"npm-codex-mcp-server__cap_2","uri":"capability://tool.use.integration.mcp.tool.schema.definition.and.registration","name":"mcp tool schema definition and registration","description":"Defines and registers MCP-compliant tool schemas that expose Codex capabilities to MCP clients. Converts Codex CLI parameters into structured MCP tool definitions with JSON schema validation, enabling Claude and other clients to discover and invoke Codex through standard tool-calling mechanisms.","intents":["I want Claude to discover what Codex can do through MCP tool introspection","I need to define input validation schemas so Claude sends properly formatted Codex requests","I want to expose Codex parameters as typed MCP tool arguments"],"best_for":["developers integrating CLI tools into MCP ecosystems","teams building Claude-compatible tool servers"],"limitations":["Schema definitions are static — no dynamic schema generation based on Codex CLI version or capabilities","Limited to MCP tool schema expressiveness — complex Codex CLI behaviors may not map cleanly to tool parameters","No schema versioning — breaking changes to Codex CLI require manual schema updates"],"requires":["MCP server framework (likely mcp npm package)","Knowledge of Codex CLI parameters and their types","JSON schema familiarity for tool definition"],"input_types":["tool parameter definitions","JSON schema specifications"],"output_types":["MCP tool schema objects","tool registration metadata"],"categories":["tool-use-integration"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"npm-codex-mcp-server__cap_3","uri":"capability://tool.use.integration.request.response.mapping.between.mcp.protocol.and.codex.cli","name":"request-response mapping between mcp protocol and codex cli","description":"Translates incoming MCP tool call requests into Codex CLI command invocations, then maps Codex CLI responses back into MCP-compliant tool result objects. Handles parameter transformation, error code mapping, and response formatting to maintain protocol compatibility across the integration boundary.","intents":["I want MCP tool calls to be transparently converted to Codex CLI commands","I need Codex CLI output formatted as proper MCP tool results","I want errors from Codex to be reported as MCP errors, not raw CLI exceptions"],"best_for":["MCP server developers bridging CLI tools and protocol clients","teams building adapter layers between incompatible tool interfaces"],"limitations":["Mapping logic is hardcoded — adding new Codex parameters requires code changes, not configuration","No bidirectional mapping validation — mismatches between MCP schema and CLI behavior can cause silent failures","Error messages from Codex may not map cleanly to MCP error codes — requires custom error translation logic"],"requires":["MCP protocol knowledge (tool call structure, result format)","Codex CLI interface documentation","Node.js async/await or Promise handling"],"input_types":["MCP tool call objects with parameters","Codex CLI stdout/stderr output"],"output_types":["MCP tool result objects","MCP error responses"],"categories":["tool-use-integration"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"npm-codex-mcp-server__cap_4","uri":"capability://safety.moderation.openai.codex.api.authentication.and.credential.management","name":"openai codex api authentication and credential management","description":"Manages OpenAI API credentials for Codex CLI authentication, reading from environment variables or configuration files and passing them to Codex CLI subprocess invocations. Ensures secure credential handling without exposing keys in logs or MCP responses.","intents":["I want the MCP server to authenticate with OpenAI Codex without hardcoding API keys","I need credentials to be passed securely to Codex CLI subprocesses","I want to support multiple authentication methods (env vars, config files)"],"best_for":["developers deploying MCP servers in production environments","teams managing API credentials across multiple integration points"],"limitations":["Credential management relies on environment variables or local files — no integration with secret management systems (AWS Secrets Manager, HashiCorp Vault, etc.)","No credential rotation or expiration handling — requires manual key updates","Credentials passed to subprocess may be visible in process listings on some systems","No audit logging of credential access or API calls"],"requires":["OpenAI API key with Codex access","Environment variable or file-based credential storage","Proper file permissions on credential files (if using file-based auth)"],"input_types":["environment variables","configuration files"],"output_types":["authenticated subprocess environment","credential validation status"],"categories":["safety-moderation","tool-use-integration"],"confidence":0.5,"matches":0,"success_rate":0}],"trust":{"score":35,"verified":false,"data_access_risk":"high","permissions":["OpenAI Codex CLI installed and in PATH","OpenAI API key configured for Codex CLI authentication","Node.js 14+ (typical MCP server runtime requirement)","MCP-compatible client (Claude desktop, custom MCP client, etc.)","Node.js child_process module (built-in)","OpenAI Codex CLI binary accessible in system PATH","Sufficient system resources for subprocess spawning","MCP server framework (likely mcp npm package)","Knowledge of Codex CLI parameters and their types","JSON schema familiarity for tool definition"],"failure_modes":["Depends on OpenAI Codex CLI being installed and authenticated locally — no fallback if CLI unavailable","MCP server runs as a single process — no built-in clustering or load balancing for concurrent requests","Latency includes CLI subprocess overhead (process spawn + IPC) on top of Codex API latency","No request queuing or rate limiting — relies on upstream MCP client to manage concurrency","Subprocess spawning adds 50-200ms overhead per invocation depending on system load","No built-in timeout handling — long-running Codex operations can block MCP responses","Requires shell/OS-level process management — behavior may differ across Linux, macOS, Windows","No streaming support — entire Codex output must be buffered before returning to MCP client","Schema definitions are static — no dynamic schema generation based on Codex CLI version or capabilities","Limited to MCP tool schema expressiveness — complex Codex CLI behaviors may not map cleanly to tool parameters","builder identity is not verified yet","no observed match outcomes yet"],"rank_breakdown":{"adoption":0.3152221788241679,"quality":0.2,"ecosystem":0.5800000000000001,"match_graph":0.25,"freshness":0.6,"weights":{"adoption":0.25,"quality":0.25,"ecosystem":0.15,"match_graph":0.23,"freshness":0.12}},"observed_outcomes":{"matches":0,"success_rate":0,"avg_confidence":0,"top_intents":[],"last_matched_at":null},"maintenance":{"status":"active","updated_at":"2026-05-24T12:16:23.901Z","last_scraped_at":"2026-05-03T14:04:47.472Z","last_commit":null},"community":{"stars":null,"forks":null,"weekly_downloads":1884,"model_downloads":null,"model_likes":null}},"distribution":{"claim_url":"https://unfragile.ai/submit?claim=codex-mcp-server","compare_url":"https://unfragile.ai/compare?artifact=codex-mcp-server"}},"signature":"ZWpUhxRPkpPiRqcBw7PzbKrewCtR6tIbr0b1z/SFYzLkv3hKzkKDnwSg7O74NarAM05kvCTwsZsOFQc7rqSCAg==","signedAt":"2026-06-21T03:08:20.332Z","signedBy":"unfragile.ai","version":1},"_links":{"self":"https://unfragile.ai/api/v1/passport/codex-mcp-server","artifact":"https://unfragile.ai/codex-mcp-server","verify":"https://unfragile.ai/api/v1/verify?slug=codex-mcp-server","publicKey":"https://unfragile.ai/api/v1/trust-passport-public-key","spec":"https://unfragile.ai/trust","schema":"https://unfragile.ai/schema.json","docs":"https://unfragile.ai/docs"}}