Development x402 offerings

OpenAI-compatible chat completions on Moonshot Kimi K2.7 Code, a coding and software-agent model with tool calling (262K context). Set your OpenAI SDK base URL to https://agents.citiviewgroup.com/v1/kimi-code and pay per token with x402 instead of an API key: $1.9 per million input and $8 per million output tokens, minimum $0.002 a call, at most $0.1742. The 402 quotes your request's maximum; "upto" payers are charged only the tokens used. Same request/response shape as OpenA

Text embeddings for semantic search and RAG (BGE-M3, multilingual, 1024 dimensions). Up to 100 texts per call, one flat price. Priced by size: $0.003 + 3x the model token cost. Also callable as GET with the same fields as query parameters.

Page action map for browser automation: every form (fields, labels, required flags), button and link on a page with ready CSS selectors, so an agent can plan clicks and form fills before opening a browser.

Find the best GitHub repositories for anything: keyword search with language, topic, minimum stars, created or pushed after filters, sorted by stars or recent activity; returns stars, forks, license, topics and last push.

Search code across all public GitHub repositories: find real usage examples of a function, API or config key, with file paths, links and the matching code fragments. Filter by language, repo, org or filename.

GitHub repository report: stars, forks, issues, license, language, activity, latest releases, recent commits and security advisories for any public repo.

OpenAI-compatible chat completions on Alibaba QwQ 32B, step-by-step reasoning for math and logic. Set your OpenAI SDK base URL to https://agents.citiviewgroup.com/v1/qwq and pay per token with x402 instead of an API key: $1.32 per million input and $2 per million output tokens, minimum $0.002 a call, at most $0.0335. The 402 quotes your request's maximum; "upto" payers are charged only the tokens used. Same request/response shape as OpenAI (reasoning, streaming supported), up

Classifier Regression Desk: Offline Paired Evaluation of JSONL Predictions. An original, dependency-free Python kit for developers comparing classifier predictions before and after a model o First approved purchase: 0.01 USDC total where maker terms permit; later 2 USDC plus tax. Stored ZIP download; commercial use, no redistribution. Verified by Speedbot for this version. Generate a private checkout_key and request_id; email and delivery consent required.

Claude Code Memory Review: CLAUDE.md, Rules, Imports — Reviews a pasted Claude Code memory setup (a folder tree plus CLAUDE.md, CLAUDE.local.md, files under .claude/rules/, imported files, AGENTS.md and an auto-memory MEMORY.md) for instructions that silently never reach the model, that load earlier or later than the author thinks, or that contradict each other. It flags an @import whose target is not in the tree (Claude Code gives no warning for it), a relative import resolved from the wrong f…

Claude Code Settings Review: Hooks, Permissions, MCP — Reviews a pasted Claude Code configuration (settings.json, hooks, permission rules, MCP server switches, scripts that start Claude Code) for keys and meanings that are easy to get wrong and that fail without an error. It flags disabledMcpjsonServers used on a server that does not come from a project's .mcp.json, a settings file with a JSON syntax error, a blocking hook that exits 1 instead of 2, the deprecated top-level approve or block dec…

Generates code comments and example snippets for your code.

AI-powered calculator that estimates technical debt from code complexity metrics, quantifies risk, and suggests remediation priorities.

Breaks code into structured components and simpler pieces for analysis or reuse.

Execute Python or JavaScript code in a secure, isolated cloud sandbox. Returns stdout, stderr, and the expression result. No setup needed; pay per call via x402. 30s timeout, 10k char code limit. A run that hits the time limit is returned with timedOut: true and the output it printed, and is charged.

Execute Python or JavaScript code in a secure, isolated cloud sandbox (Pro tier). Same as /api/code-run but with 60s timeout and 50k char code limit for longer computations. Returns stdout, stderr, and the expression result. A run that hits the time limit is returned with timedOut: true and the output it printed, and is charged.

Run code in an isolated cloud sandbox and get its output — agents can't self-host this. POST {code, language?}. language: python (default), javascript, bash, r, java. Returns stdout, stderr, results, error. Hard-capped: 25s wall-clock, 100KB code, 64KB output, fully isolated.

AI-powered analyzer that maps safe upgrade and migration paths for codebases or smart contracts, highlighting risks, dependencies, and steps.

Detects code smells, rates their severity, and provides targeted refactoring suggestions to improve code quality.

Render Markdown (GitHub-flavored: tables, fenced code, task lists) to clean HTML with raw inline HTML stripped. POST { markdown } (max 500KB).

AI-powered generator that produces ready-to-use design system UI components, code snippets, and tokens from prompts to accelerate product design.

Run the same input through gzip, brotli, and raw deflate at max settings and report each one's output size and ratio. Use this to decide which algorithm to actually use - the answer depends on your content (binary often resists compression entirely; English prose squeezes well with brotli; JSON sits in between).

Generate a single-file HTML page from a text brief

Validate, pretty-print (0-8 spaces), or minify JSON — json formatter/validator.

[MCP: optimize_context] Pay-per-call JSON context pruning & LLM token budget enforcement. Saves up to 85% prompt tokens.