Model Context Protocol (MCP) server for Citadel Predict pre-execution AI agent cost estimation
Wellknown found it in public sources; nobody has proven control of it yet. Claiming takes one click if the repository is under your GitHub account, or a small file on your domain otherwise. Verified owners get the badge, 15-minute checks, status alerts, edits that outrank crawled data, and a ranking boost.
Agents can do it too: POST https://wellknown.network/api/v1/claims with {"agent":"citadel-predict-mcp","method":"well_known_file"} — machine-readable steps at claim.json, guide at /docs/claim.
Everything here was measured by our prober or read from a registry. Nothing is self-reported.
Attributed to the source that supplied each field. Treated as claims, not facts.
# citadel-predict-mcp [](https://pypi.org/project/citadel-predict-mcp/) [](https://opensource.org/licenses/MIT) **Model Context Protocol (MCP) server for Citadel Predict pre-execution AI agent cost estimation.** `citadel-predict-mcp` connects your hosted [Citadel Predict API](https://github.com/Athullvr/Citadel) directly into **Claude Desktop** and **Claude Code** via a local standard I/O (stdio) MCP server. With this server configured, Claude can natively estimate token usage ranges ($low, expected, high$) and flag out-of-distribution risks for autonomous workflows **before running them**—without requiring manual CLI execution. --- ## Features - ⚡ **Native Claude Tool Calling**: Claude automatically decides when to call `estimate_agent_cost` when planning or dispatching tasks. - 🔒 **Zero Network Exposure**: Runs strictly as a local `stdio` subprocess spawned by Claude Desktop / Claude Code. - 🎯 **Pre-Execution Guardrails**: Predicts token consumption bounds before multi-step tools or reasoning loops execute. - 🛡️ **User-Friendly Error Handling**: Catches authentication, rate-limiting, and validation issues, presenting clear, actionable suggestions to Claude rather than raw stack traces. --- ## Installation Install the package via `pip`: ```bash pip install citadel-predict-mcp ``` *(For local development from the repository root: `pip install -e packages/citadel-predict-mcp`)* Verifying installation: ```bash citadel-predict-mcp --help ``` --- ## Quickstart: One-Command Auto-Configuration (Recommended) To automatically configure **both** Claude Desktop and Claude Code on Windows, macOS, or Linux with zero manual JSON editing: ```bash # Run from repository root with your active Python environment: python scripts/configure_mcp.py # Or pass parameters non-interactively: python scripts/configure_mcp.py -…
Mapped onto the structured taxonomy from declared text and observed tool names. Confidence shown for derived entries.
Every source is kept verbatim. Field changes are logged as events.