Diogenes Autonomous Microservice Gateway

Evaluates untrusted agent prompts, bash commands, and tool arguments to block prompt injection, destructive shell commands, and private-key exfiltration. Use when screening a general shell command or user prompt before agent execution; for statically validating SQL database queries at the AST level, use validate_sql_query_ast instead. Stateless read-only edge call (max 32 KB, 60 RPM): accepts 'payload' ('action_type' defaults to 'bash_command'), supports '?demo=true' or L402 (100 sats), and returns HTTP 400 if invalid.

Performs static AST syntax tree parsing on SQL statements to verify strictly read-only SELECT clauses and reject mutations, stacked queries, or comment bypasses. Use when verifying that an agent-authored SQL statement is strictly read-only before running it on a database; for screening non-SQL bash commands or general agent prompts, use verify_agent_command_safety instead. Stateless read-only edge call (max 32 KB, 60 RPM): accepts 'sql_statement' ('dialect' defaults to 'postgres'), supports '?demo=true' or L402 (75 sats), and returns HTTP 400 if invalid.

Audits Python source code for forbidden syscalls (os, subprocess, socket, open, eval) inside an AST-governed WebAssembly sandbox with no host filesystem access. Use when executing LLM-generated Python snippets safely; for checking shell commands without executing code, use verify_agent_command_safety instead. Stateless read-only edge call (max 32 KB, 60 RPM): accepts 'python_script' ('timeout_profile' defaults to 'fast_1s'), supports '?demo=true' or MPP (1500 sats), and returns HTTP 400 if invalid.

Scans an external MCP server's initial tools/list manifest to detect tool-name shadowing (read_file, bash) and hidden prompt-injection directives in descriptions. Use when vetting a newly discovered third-party MCP server before mounting it into an agent session; for comparing a previously pinned MCP baseline against a live mid-session tools/list response, use detect_mcp_schema_rugpull instead. Stateless read-only edge call (max 32 KB, 60 RPM): accepts 'manifest_json' ('audit_profile' defaults to 'strict'), supports '?demo=true' or L402 (150 sats), and returns HTTP 400 if invalid.

Computes structural diffs between a pinned baseline MCP tools/list snapshot and a live server response to catch mid-session description or schema tampering. Use when verifying that a connected MCP server has not mutated its tool definitions mid-session; for performing a first-time static security audit on a single MCP manifest, use audit_mcp_server_manifest instead. Stateless read-only edge call (max 32 KB, 60 RPM): accepts 'baseline_tools_json' ('current_tools_json' defaults to 'standard'), supports '?demo=true' or L402 (85 sats), and returns HTTP 400 if invalid.

Audits inbound MCP sampling/createMessage callbacks from connected servers to block cross-server context snooping (includeContext='allServers') and token quota hijacking. Use when filtering a connected MCP server's request to sample the host LLM; for scrubbing text returned from a standard synchronous tools/call response, use sanitize_tool_output_stream instead. Stateless read-only edge call (max 32 KB, 60 RPM): accepts 'sampling_request_json' ('max_tokens_ceiling' defaults to '512'), supports '?demo=true' or L402 (100 sats), and returns HTTP 400 if invalid.

Sanitizes untrusted external tool or web-scrape outputs by stripping invisible Unicode Tag characters (U+E0000), neutralizing Markdown tracking image beacons, and defusing indirect prompt overrides. Use when cleaning raw third-party tool outputs or scraped web text before inserting it into the LLM context window; for redacting personal identifiable information (emails, SSNs, keys) from outbound text, use redact_sensitive_pii instead. Read-only edge call (32 KB, 60 RPM): accepts 'raw_tool_output' (with enum defaults), uses '?demo=true' or L402 (75 sats); HTTP 400 if invalid.

Redacts emails, phone numbers, SSNs, credit cards, API keys, and crypto private keys from text before logging or external transmission. Use when stripping personal data or secret credentials from prompts, logs, or outbound messages; for neutralizing indirect prompt injections and invisible Unicode in inbound tool responses, use sanitize_tool_output_stream instead. Stateless read-only edge call (max 32 KB, 60 RPM): accepts 'text' ('redaction_style' defaults to 'typed_token'), supports '?demo=true' or L402 (150 sats), and returns HTTP 400 if invalid.

Analyzes an agent's recent tool-call trace to detect identical retry spins, A/B state oscillation, and runaway token burn. Use when monitoring multi-step agent execution loops to decide whether to continue or halt and escalate; for inspecting a single command payload for security threats before execution, use verify_agent_command_safety instead. Stateless read-only edge call (max 32 KB, 60 RPM): accepts 'action_trace_json' ('sensitivity' defaults to 'strict_2_repeats'), supports '?demo=true' or L402 (70 sats), and returns HTTP 400 if invalid.

Verifies a remote agent's A2A /.well-known/agent.json discovery card, checking JWS/DID signature binding, endpoint HTTPS enforcement, and capability spoofing. Use when authenticating a peer autonomous agent before delegating subtasks or exchanging payments over A2A; for auditing an MCP tool server's tools/list manifest rather than a peer agent's A2A card, use audit_mcp_server_manifest instead. Read-only edge call (32 KB, 60 RPM): accepts 'agent_card_json' ('trust_policy' defaults to 'require_https_and_did'), supports '?demo=true' or L402 (90 sats), and returns HTTP 400 if invalid.

Static validation firewall for outbound network destination URIs against cloud metadata IP addresses (169.254.169.254), RFC 1918 private IP blocks, and query-string secret exfiltration before agent dispatch. Use when validating a user-supplied or tool-supplied URI destination before network dispatch; for auditing local shell commands or general prompts, use verify_agent_command_safety instead. Read-only edge call (32 KB, 60 RPM): accepts 'destination_uri' ('egress_policy' defaults to 'block_private_and_metadata'), supports '?demo=true' or L402 (65 sats), and returns HTTP 400 if invalid.

Cross-references factual claims, numbers, and citations in a draft agent response against retrieved source passages to flag unsupported hallucinations before delivery. Use when verifying that a generated answer is 100% grounded in source evidence before returning it to the user; for scrubbing inbound tool outputs for prompt-injection and invisible Unicode, use sanitize_tool_output_stream instead. Stateless read-only edge call (max 32 KB, 60 RPM): accepts 'claim' ('context' defaults to 'standard'), supports '?demo=true' or L402 (150 sats), and returns HTTP 400 if invalid.

Statically audits Solidity, Vyper, or Move smart contract source code for honeypot transfer taxes, hidden mint backdoors, proxy upgrade traps, and reentrancy. Use when auditing smart contract source code or bytecode before deploying or interacting with a protocol; for scanning Solana Anchor/Rust programs specifically for Cross-Program Invocation (CPI) signer/account vulnerabilities, use scan_solana_cpi_reentrancy instead. Read-only edge call (32 KB, 60 RPM): accepts 'code_or_address' ('chain' defaults to 'evm_solidity'), uses '?demo=true' or L402 (250 sats); HTTP 400 if invalid.

Scans Solana Anchor and native Rust program instructions for unverified Cross-Program Invocation (CPI) program IDs, missing signer checks, and PDA bump seed canonicalization bugs. Use when auditing Solana Rust/Anchor source code or instruction account structs for CPI privilege escalation; for checking live SPL token mint/freeze authorities and rug-pull risk by mint address, use evaluate_solana_token_risk instead. Read-only edge call (32 KB, 60 RPM): accepts 'rust_source_or_idl' ('framework' defaults to 'anchor'), uses '?demo=true' or L402 (160 sats); HTTP 400 if invalid.

Evaluates a Solana SPL token's rug-pull probability by checking active mint authority, freeze authority, mutable metadata, and top-holder concentration. Use when making a pre-trade security decision on whether a Solana SPL token is safe to buy; for retrieving SPL token symbol, decimals, and DEX liquidity pool metadata via x402, use inspect_solana_token_liquidity instead. Stateless read-only edge call (max 32 KB, 60 RPM): accepts 'token_address' ('risk_profile' defaults to 'strict_decentralized'), supports '?demo=true' or L402 (100 sats), and returns HTTP 400 if invalid.

Retrieves canonical Solana SPL token metadata including name, ticker symbol, decimal precision, and current mint/freeze authority addresses. Use when resolving SPL token decimals, ticker symbols, and registry metadata for portfolio or swap formatting; for computing a composite security risk score and binary should_transact verdict, use evaluate_solana_token_risk instead. Stateless read-only edge call (max 32 KB, 60 RPM): accepts 'mint' ('network' defaults to 'mainnet-beta'), supports '?demo=true' or x402 (200 sats), and returns HTTP 400 if invalid.

Tracks 24-hour net accumulation or distribution flows across top Solana whale wallets and flags coordinated insider wallet clusters for a given SPL token. Use when analyzing on-chain smart-money order flow and holder clustering on Solana; for checking static SPL contract mint/freeze authority permissions, use evaluate_solana_token_risk instead. Stateless read-only edge call (max 32 KB, 60 RPM): accepts 'token' ('window' defaults to '1h'), supports '?demo=true' or x402 (500 sats), and returns HTTP 400 if invalid.

Calculates optimal Solana compute-unit (CU) microlamport priority fees and Jito bundle tip lamports to achieve target block-inclusion probability. Use when setting SetComputeUnitPrice and Jito tips before broadcasting a Solana transaction; for querying Bitcoin L1 sat/vB mempool fee rates, use estimate_bitcoin_mempool_fees instead. Stateless read-only edge call (max 32 KB, 60 RPM): accepts 'priority_level' ('tx_type' defaults to 'simple_transfer'), supports '?demo=true' or x402 (50 sats), and returns HTTP 400 if invalid.

Decodes EIP-712 and Uniswap Permit2 off-chain typed-data payloads to verify spender identity, detect unlimited (uint256.max) token approvals, and flag expired deadlines. Use when inspecting an off-chain eth_signTypedData_v4 or Permit2 payload before signing; for decoding raw on-chain hex transaction calldata (0x...) for eth_sendTransaction, use simulate_evm_transaction_calldata instead. Stateless read-only edge call (max 32 KB, 60 RPM): accepts 'typed_data_json' ('chain_id' defaults to '1'), supports '?demo=true' or L402 (100 sats), and returns HTTP 400 if invalid.

Decodes raw EVM hexadecimal transaction calldata (0x...) by 4-byte function selector to expose hidden ERC-20 approve/transferFrom calls, setApprovalForAll drains, and multicall bundles. Use when auditing raw hex transaction data (to, value, data) before broadcasting an on-chain EVM transaction; for inspecting off-chain EIP-712 JSON typed-data signatures, use decode_eip712_permit_signature instead. Stateless read-only edge call (max 32 KB, 60 RPM): accepts 'calldata_hex' ('network' defaults to 'base_mainnet'), supports '?demo=true' or L402 (90 sats), and returns HTTP 400 if invalid.

Evaluates decentralized exchange (DEX) swap parameters against pool depth and mempool visibility to quantify MEV sandwich risk and compute optimal slippage tolerance in basis points. Use when configuring slippage tolerance and private-RPC routing before executing an on-chain DEX token swap; for comparing cross-chain bridge settlement routes across different L1 blockchains, use compare_crosschain_bridge_routes instead. Read-only edge call (32 KB, 60 RPM): accepts 'pair' ('execution_venue' defaults to 'uniswap_v3_public'), uses '?demo=true' or L402 (150 sats); HTTP 400 if invalid.

Compares a target recipient wallet address against recent transaction history to detect vanity prefix/suffix lookalike spoofing and zero-value dusting attacks. Use when verifying a recipient EVM or Solana wallet address before sending funds; for decoding hex smart contract calldata payloads, use simulate_evm_transaction_calldata instead. Stateless read-only edge call (max 32 KB, 60 RPM): accepts 'target_address' ('address_standard' defaults to 'auto_detect'), supports '?demo=true' or L402 (100 sats), and returns HTTP 400 if invalid.

Decodes inbound HTTP 402 Payment Required challenges (L402 BOLT11 + Macaroon, x402 Solana Native SOL, MPP Base intents) to verify cryptographic payment-hash binding and block overcharge fraud. Use when auditing a 402 challenge header before an autonomous agent pays a Lightning invoice or USDC transfer; for checking on-chain Bitcoin L1 mempool feerates, use estimate_bitcoin_mempool_fees instead. Stateless read-only edge call (max 32 KB, 60 RPM): accepts 'www_authenticate' ('protocol_hint' defaults to 'L402'), supports '?demo=true' or L402 (60 sats), and returns HTTP 400 if invalid.

Fetches real-time Bitcoin mainnet mempool feerate recommendations in satoshis per virtual byte (sat/vB) across fastest, half-hour, hour, and economy confirmation tiers. Use when checking current Bitcoin L1 network congestion before broadcasting an on-chain BTC transaction; for calculating exact SegWit/Taproot UTXO batch consolidation vByte weights and dust thresholds, use plan_bitcoin_utxo_consolidation instead. Stateless read-only edge call (max 32 KB, 60 RPM): accepts 'confirmation_target', supports '?demo=true' or L402 (50 sats), and returns HTTP 400 if invalid.

Analyzes a set of Bitcoin UTXO balances in satoshis to compute exact vByte transaction weight, identify uneconomical dust inputs, and recommend whether to consolidate now or hold. Use when optimizing on-chain Bitcoin treasury UTXO sets to minimize long-term miner fee drag; for fetching current live Bitcoin mempool sat/vB feerates alone, use estimate_bitcoin_mempool_fees instead. Stateless read-only edge call (max 32 KB, 60 RPM): accepts 'utxos_sats_json' ('script_type' defaults to 'p2wpkh_segwit'), supports '?demo=true' or L402 (80 sats), and returns HTTP 400 if invalid.

Calculates instantaneous spot exchange rates and equivalent USD, BTC, and satoshi values across BTC, ETH, SOL, SUI, and USDC. Use when converting a token amount into USD or satoshis for invoice pricing or accounting; for quoting executable non-KYC cross-chain bridge routes with bridge fees and settlement times, use compare_crosschain_bridge_routes instead. Stateless read-only edge call (max 32 KB, 60 RPM): accepts 'currency' ('amount' defaults to 'standard'), supports '?demo=true' or L402 (100 sats), and returns HTTP 400 if invalid.

Compares non-custodial, non-KYC cross-chain swap routes (THORChain native L1 swaps, Maya Protocol, Blockstream Liquid) by basis-point fee, net output sats, and settlement time. Use when routing cross-chain treasury settlements (e.g. Solana Native SOL or Base ETH into native Bitcoin); for calculating mid-market spot exchange rates without bridge routing fees, use calculate_spot_swap_rate instead. Stateless read-only edge call (max 32 KB, 60 RPM): accepts 'from_asset' ('to_asset' defaults to 'BTC_NATIVE'), supports '?demo=true' or L402 (100 sats), and returns HTTP 400 if invalid.

Generates text completions and structured reasoning across all 45 Google Gemini foundation models (Gemini 4 Argon, 3.1 Pro, 3.8 Flash) with zero Google account or API key required. Use when running single-turn reasoning, coding, or analytical prompts on a specific Gemini model; for delegating multi-step autonomous tasks that require remote tool execution, use execute_antigravity_agent instead. Stateless read-only edge call (max 32 KB, 60 RPM): accepts 'prompt' ('model' defaults to 'gemini-4-argon'), supports '?demo=true' or L402 (250 sats), and returns HTTP 400 if invalid.

Dispatches a complex engineering or operational task to Google's Antigravity autonomous agent for multi-step planning, tool orchestration, and verified execution. Use when solving multi-stage software engineering or workflow tasks that require autonomous planning; for compiling exhaustive multi-source literature and market research dossiers, use conduct_deep_research_investigation instead. Stateless read-only edge call (max 32 KB, 60 RPM): accepts 'input' ('execution_mode' defaults to 'autonomous_verified'), supports '?demo=true' or L402 (500 sats), and returns HTTP 400 if invalid.

Runs an iterative, multi-hop investigative research session via Gemini Deep Research Pro, cross-examining primary sources to produce a cited technical report. Use when investigating complex scientific, architectural, or market questions requiring deep multi-hop synthesis; for running a fast single-turn Google Search grounded fact lookup, use search_gemini_grounded_web instead. Stateless read-only edge call (max 32 KB, 60 RPM): accepts 'input' ('depth_tier' defaults to 'quick_synthesis'), supports '?demo=true' or L402 (1000 sats), and returns HTTP 400 if invalid.

Plans and executes OS-level desktop, browser DOM, and GUI coordinate actions (click, type, scroll, wait) using the Gemini Computer Use agent. Use when automating interactive web UIs or desktop workflows that lack a direct REST API; for running code-first API and terminal automation tasks, use execute_antigravity_agent instead. Stateless read-only edge call (max 32 KB, 60 RPM): accepts 'input' ('viewport_preset' defaults to 'desktop_1080p'), supports '?demo=true' or L402 (800 sats), and returns HTTP 400 if invalid.

Answers factual queries with live Google Search grounding, returning verified up-to-date answers accompanied by source URLs and grounding metadata. Use when verifying live web facts, documentation updates, or current events in a single low-latency call; for querying geographic places, coordinates, directions, and local business data via Google Maps grounding, use search_gemini_grounded_maps instead. Stateless read-only edge call (max 32 KB, 60 RPM): accepts 'prompt' ('recency_bias' defaults to 'past_24h'), supports '?demo=true' or L402 (200 sats), and returns HTTP 400 if invalid.

Executes geospatial and point-of-interest queries grounded in live Google Maps data, returning verified addresses, coordinates, operating hours, and spatial context. Use when resolving physical locations, venues, transit distances, or regional infrastructure data; for searching general web articles, documentation, and news, use search_gemini_grounded_web instead. Stateless read-only edge call (max 32 KB, 60 RPM): accepts 'prompt' ('distance_units' defaults to 'metric_km'), supports '?demo=true' or L402 (250 sats), and returns HTTP 400 if invalid.

Compiles a structured JSON research dossier ('key_findings', 'synthesis_quality_score', 'sources_cited') settled via Base EVM Machine Payments Protocol (MPP). Use when generating a strict-JSON cited intelligence brief paid via Base Native ETH MPP session headers; for running long-form narrative research reports via L402 Lightning, use conduct_deep_research_investigation instead. Stateless read-only edge call (max 32 KB, 60 RPM): accepts 'research_topic' ('output_density' defaults to 'concise_5_points'), supports '?demo=true' or MPP (1000 sats), and returns HTTP 400 if invalid.

Generates normalized 3,072-dimensional or 768-dimensional semantic vector embeddings via Google Gemini Embedding 2 for RAG indexing and cosine-similarity search. Use when embedding text passages or queries via L402 Lightning with configurable Matryoshka dimensions; for running live Google Search grounded factual queries, use search_gemini_grounded_web instead. Stateless read-only edge call (max 32 KB, 60 RPM): accepts 'text' ('model' defaults to 'gemini-embedding-2'), supports '?demo=true' or L402 (50 sats), and returns HTTP 400 if invalid.

Analyzes combined multimodal prompts spanning text, image descriptors, audio transcripts, and video temporal events using Gemini Omni Flash. Use when correlating information across multiple media modalities in a single reasoning pass; for running pure text-only reasoning or coding prompts, use generate_gemini_completion instead. Stateless read-only edge call (max 32 KB, 60 RPM): accepts 'input' ('modality_focus' defaults to 'cross_modal_synthesis'), supports '?demo=true' or L402 (300 sats), and returns HTTP 400 if invalid.

Computes 3D spatial affordances, obstacle-avoidance waypoints, and manipulator grasp sequences using Gemini Robotics ER 2 embodied reasoning. Use when planning physical robot arm kinematics, mobile navigation waypoints, or spatial scene task orders; for interacting with 2D desktop or browser GUI screens, use operate_computer_use_agent instead. Stateless read-only edge call (max 32 KB, 60 RPM): accepts 'prompt' ('coordinate_frame' defaults to 'cartesian_xyz_meters'), supports '?demo=true' or L402 (450 sats), and returns HTTP 400 if invalid.

Generates photorealistic illustrations, UI mockups, and technical diagrams from text prompts using the Nano Banana Pro (Gemini 2.5 Flash Image) model family. Use when creating static visual assets, icons, or architecture diagrams from text descriptions; for synthesizing animated multi-second video clips with motion, use generate_veo_cinematic_video instead. Stateless read-only edge call (max 32 KB, 60 RPM): accepts 'prompt' ('aspect_ratio' defaults to '1:1'), supports '?demo=true' or L402 (350 sats), and returns HTTP 400 if invalid.

Synthesizes high-definition cinematic video sequences with temporally coherent camera motion and physics using Google Veo 3. Use when generating moving video clips, product B-roll, or animated scene renders from text prompts; for rendering static single-frame images or diagrams at lower cost, use generate_nano_banana_image instead. Stateless read-only edge call (max 32 KB, 60 RPM): accepts 'prompt' ('motion_style' defaults to 'static_tripod'), supports '?demo=true' or L402 (1500 sats), and returns HTTP 400 if invalid.

Composes studio-grade instrumental music tracks, ambient loops, and multi-instrument arrangements from genre, tempo, and mood prompts using Google Lyria 3. Use when generating musical scores, soundtrack stems, or rhythmic audio loops; for converting written text into spoken human voice narration (TTS), use synthesize_gemini_speech_audio instead. Stateless read-only edge call (max 32 KB, 60 RPM): accepts 'prompt' ('genre_preset' defaults to 'cinematic_synth'), supports '?demo=true' or L402 (500 sats), and returns HTTP 400 if invalid.

Converts written text into natural, prosody-controlled spoken voice audio using Gemini 3.8 Flash Neural Text-to-Speech. Use when rendering voiceover narration or spoken agent replies from a deterministic text script; for conducting bidirectional conversational audio dialog comprehension, use process_live_audio_dialogue instead. Stateless read-only edge call (max 32 KB, 60 RPM): accepts 'text' ('voice_persona' defaults to 'neutral_broadcast'), supports '?demo=true' or L402 (200 sats), and returns HTTP 400 if invalid.

Processes real-time conversational dialog turns with native acoustic emotion, interruption handling, and paralinguistic cue comprehension via Gemini 3.8 Live. Use when handling interactive voice-agent dialog turns that require tone and turn-taking awareness; for performing one-way text-to-speech readout of a static text string, use synthesize_gemini_speech_audio instead. Stateless read-only edge call (max 32 KB, 60 RPM): accepts 'input' ('response_tone' defaults to 'direct_peer'), supports '?demo=true' or L402 (300 sats), and returns HTTP 400 if invalid.

Compresses large source files, logs, or transcripts by 50–80% to fit an LLM context budget while preserving 100% of function/class/type signatures, imports, error frames, and query-matched lines. Use when shrinking oversized source files or logs before feeding them into an LLM prompt; for counting BPE tokens across prompt segments and calculating multi-model inference costs without altering text, use plan_context_token_budget instead. Read-only edge call (32 KB, 60 RPM): accepts 'content' (with enum defaults), uses '?demo=true' or L402 (50 sats); HTTP 400 if invalid.

Counts subword BPE tokens across prompt segments, detects high-priority instructions trapped in the 35%–72% Lost-in-the-Middle attention trough, and calculates USD/satoshi costs across frontier LLMs. Use when auditing prompt token budgets, reordering prompt sections for primacy/recency attention, or comparing model API costs; for actively compressing and stripping low-salience lines from a single large file, use prune_context_skeleton instead. Read-only edge call (32 KB, 60 RPM): accepts 'segments_json' (with enum defaults), uses '?demo=true' or L402 (30 sats); HTTP 400 if invalid.

Reranks and deduplicates candidate RAG or retrieval passages against a search prompt using Okapi BM25 (k1=1.5, b=0.75), Reciprocal Rank Fusion (RRF), and Maximal Marginal Relevance (MMR) Jaccard diversity filtering. Use when selecting the top-K highest-signal, non-redundant passages from vector search or grep results; for compressing a single contiguous source file while keeping its original line order, use prune_context_skeleton instead. Read-only edge call (32 KB, 60 RPM): accepts 'search_prompt' ('documents_json' defaults to 'standard'), uses '?demo=true' or L402 (50 sats); HTTP 400 if invalid.

Applies unified diffs (@@ ... @@) or SEARCH/REPLACE blocks to source code using sliding-window token similarity to resolve drifted line numbers, align indentation, and verify bracket balance. Use when healing and applying a code edit or diff that failed exact line matching due to line-number drift or whitespace differences; for inspecting function signatures, call trees, and cyclomatic complexity without modifying code, use analyze_ast_symbol_graph instead. Read-only call (32 KB, 60 RPM): accepts 'original_code' (with enum defaults), uses '?demo=true' or L402 (75 sats); HTTP 400 if invalid.

Static lexer and AST parser only (zero execution or evaluation): extracts imports, exports, classes, interfaces, and functions with line numbers, intra-file call edges, McCabe cyclomatic complexity, and Halstead volume across TypeScript, Python, Rust, Go, and Move. Use when mapping internal functions, call graphs, and complexity hotspots of a source file before refactoring; for analyzing cross-file import DAGs and circular dependencies, use audit_module_dependency_cycles instead. Read-only call (32 KB, 60 RPM): accepts 'source_text' (with enum defaults), uses '?demo=true' or L402 (75 sats); HTTP 400 if invalid.

Runs Tarjan's Strongly Connected Components (SCC) algorithm on a module import map to detect circular dependencies, compute parallel topological build waves, and trace downstream change blast radius. Use when auditing multi-file import graphs for circular cycles or determining which downstream files must be re-tested after an edit; for extracting function-level symbols and cyclomatic complexity inside a single source file, use analyze_ast_symbol_graph instead. Read-only call (32 KB, 60 RPM): accepts 'modules_json' (with enum defaults), uses '?demo=true' or L402 (75 sats); HTTP 400 if invalid.

Parses noisy CI/terminal logs and multi-language stack traces (Node/TS, Python, Rust, Go), strips node_modules/stdlib frames to isolate the innermost failing user-code frame, and clusters repetitive logs into Drain templates. Use when triage-debugging test failures, crash dumps, or repetitive server error logs; for compressing general source code files or prose transcripts, use prune_context_skeleton instead. Read-only edge call (32 KB, 60 RPM): accepts 'raw_log' ('runtime_hint' defaults to 'auto_detect'), supports '?demo=true' or L402 (50 sats), and returns HTTP 400 if invalid.

Inspects a sample JSON payload (inferring UUID, ISO-8601 datetime, email, integer vs float, and array item types) and compiles strict TypeScript interfaces, Zod schemas, Python Pydantic v2 models, and JSON Schema Draft 2020-12. Use when generating static type definitions and runtime validators in TypeScript and Python from a sample JSON response; for extracting existing function/class symbols and complexity metrics from source code, use analyze_ast_symbol_graph instead. Read-only call (32 KB, 60 RPM): accepts 'sample_json' (with enum defaults), uses '?demo=true' or L402 (60 sats); HTTP 400 if invalid.

Strict Shunting-Yard AST arithmetic parser with zero eval() or dynamic code execution: evaluates algebraic formulas and computes exact financial NPV, Newton-Raphson IRR, Kelly criterion, and dataset stats. Use when computing exact arithmetic formulas, cashflow NPV/IRR, or descriptive statistics without LLM floating-point guesswork; for running two-sample A/B hypothesis tests, use evaluate_ab_statistical_significance instead. Read-only edge call (32 KB, 60 RPM): accepts 'math_formula' ('precision_digits' defaults to '2_decimals'), uses '?demo=true' or L402 (40 sats); HTTP 400 if invalid.

Solves a task dependency DAG via the Critical Path Method (CPM) to compute ES/EF/LS/LF, Total Float/Slack, optimal makespan, and the Critical Path. Use when scheduling dependent engineering milestones or workflow steps to find the minimum project duration and zero-slack bottleneck path; for detecting circular code imports and file blast radius without task durations, use audit_module_dependency_cycles instead. Read-only edge call (32 KB, 60 RPM): accepts 'tasks_json' ('time_unit' defaults to 'minutes'), supports '?demo=true' or L402 (80 sats), and returns HTTP 400 if invalid.

Computes Welch's two-sample t-test or two-proportion z-test, returning two-tailed p-values, Cohen's d effect size, relative lift %, and Bayesian P(Treatment > Control). Use when determining whether a prompt change, model benchmark, or A/B experiment produced a statistically significant improvement; for evaluating single algebraic formulas or NPV/IRR cashflows, use evaluate_deterministic_math instead. Read-only edge call (32 KB, 60 RPM): accepts 'control_json' ('treatment_json' defaults to 'standard'), supports '?demo=true' or L402 (60 sats), and returns HTTP 400 if invalid.

Compares baseline and candidate OpenAPI 3.1 specs, JSON Schemas, or MCP tool manifests to detect breaking changes (removed fields, new required inputs, narrowed enums, type mutations). Use when auditing API, JSON Schema, or MCP contract revisions for backward compatibility before deployment; for compiling new TypeScript, Zod, and Pydantic schemas from a single sample JSON payload, use compile_polyglot_type_schemas instead. Read-only edge call (32 KB, 60 RPM): accepts 'baseline_schema_json' (with enum defaults), uses '?demo=true' or L402 (75 sats); HTTP 400 if invalid.

Synthesizes deterministic unit-test vectors from a JSON Schema or function signature covering valid equivalence partitions, numeric/length boundaries, and adversarial Unicode/null inputs. Use when generating exhaustive edge-case unit test payloads and fuzz vectors for an API endpoint or function; for healing and applying unified diffs or search/replace edits to source code, use apply_fuzzy_code_patch instead. Read-only edge call (32 KB, 60 RPM): accepts 'schema_or_signature_json' (with enum defaults), uses '?demo=true' or L402 (65 sats); HTTP 400 if invalid.

Runs a reproducible seeded Mulberry32 Monte Carlo simulation over uncertain cost, latency, or duration variables to compute P10/P50/P90/P95 percentiles, Value-at-Risk (VaR), and CVaR. Use when modeling stochastic uncertainty, tail risk (VaR/CVaR), and probability of meeting a budget or SLA target; for evaluating deterministic single-point algebraic formulas or exact NPV/IRR cashflows, use evaluate_deterministic_math instead. Read-only edge call (32 KB, 60 RPM): accepts 'variables_json' (with enum defaults), uses '?demo=true' or L402 (70 sats); HTTP 400 if invalid.

Computes structural diffs between two agent state snapshots ('added_keys', 'mutated_keys', 'removed_keys'), flags boolean belief contradictions, generates an RFC 7396 JSON Merge Patch delta, and computes a canonical checkpoint hash. Use when checkpointing agent state or transmitting minimal RFC 7396 state deltas across multi-agent handoffs; for counting BPE tokens and auditing Lost-in-the-Middle prompt attention, use plan_context_token_budget instead. Read-only call (32 KB, 60 RPM): accepts 'previous_state_json' (with enum defaults), uses '?demo=true' or L402 (50 sats); HTTP 400 if invalid.

Deterministic JSON syntax repair: normalizes quotes, unclosed brackets, and boolean literals into compliant RFC 8259 JSON structure without data deletion. Use when fixing syntax-broken JSON strings that fail JSON.parse(); for coercing field types and verifying properties on already-parseable JSON objects against a target JSON Schema, use enforce_json_schema_types instead. Read-only edge call (32 KB, 60 RPM): accepts 'malformed_json' (with enum defaults), uses '?demo=true' or L402 (50 sats); HTTP 400 if invalid.

Validates a raw LLM output object against a target JSON Schema, automatically coercing mismatched scalar types (e.g. string '42' to integer 42), injecting schema defaults, and pruning undeclared hallucinated fields. Use when enforcing schema compliance and type safety on structured LLM outputs before passing them to typed APIs; for healing unparseable raw JSON text with broken quotes or trailing commas, use heal_malformed_json instead. Read-only edge call (32 KB, 60 RPM): accepts 'raw_output' (with enum defaults), uses '?demo=true' or L402 (75 sats); HTTP 400 if invalid.

Extracts structured JSON records, named entities, dates, and financial figures from unstructured prose, emails, or HTML/Markdown dumps. Use when converting unstructured natural-language text into a structured JSON array of entities; for validating and coercing an existing JSON payload against a formal JSON Schema, use enforce_json_schema_types instead. Stateless read-only edge call (max 32 KB, 60 RPM): accepts 'text' ('extraction_target' defaults to 'general_entities_and_metrics'), supports '?demo=true' or L402 (150 sats), and returns HTTP 400 if invalid.

Deduplicates repetitive lines or records, normalizes Unicode whitespace, strips HTML/control noise, and canonicalizes tabular text or JSON arrays. Use when cleaning noisy scraped lists, CSV rows, or duplicate log lines before indexing or analysis; for extracting typed JSON fields from free-form prose paragraphs, use extract_structured_json_entities instead. Stateless read-only edge call (max 32 KB, 60 RPM): accepts 'data' ('normalization_mode' defaults to 'exact_match'), supports '?demo=true' or L402 (100 sats), and returns HTTP 400 if invalid.

Synthesizes linear-time O(n) regular expressions free of catastrophic backtracking (ReDoS) and validates them against positive and negative test strings. Use when generating and verifying a regular expression pattern for input validation or lexical parsing; for repairing broken JSON syntax directly without writing custom regexes, use heal_malformed_json instead. Stateless read-only edge call (max 32 KB, 60 RPM): accepts 'description' ('flavor' defaults to 'ecmascript_js'), supports '?demo=true' or L402 (50 sats), and returns HTTP 400 if invalid.

Compiles a raw task description into a structured XML system prompt (<role>, <context>, <objective>, <constraints>, <workflow>, <output_format>) engineered to prevent instruction drift. Use when authoring or upgrading a system prompt for an autonomous agent or sub-agent workflow; for mapping a task objective to a concrete sequence of MCP tool calls, use route_mcp_workflow_plan instead. Stateless read-only edge call (max 32 KB, 60 RPM): accepts 'task' ('prompt_style' defaults to 'anthropic_xml_tags'), supports '?demo=true' or L402 (100 sats), and returns HTTP 400 if invalid.

Synthesizes PEP 8 compliant, self-contained Python 3 scripts, data parsers, and CLI utilities complete with type hints, docstrings, and inline assertion tests. Use when generating a standalone Python automation script or parser from a functional specification; for summarizing existing documentation or codebases into executive bullet points, use summarize_technical_document instead. Stateless read-only edge call (max 32 KB, 60 RPM): accepts 'prompt' ('dependency_policy' defaults to 'stdlib_only'), supports '?demo=true' or L402 (300 sats), and returns HTTP 400 if invalid.

Condenses long technical specifications, RFCs, research papers, or meeting transcripts into high-density executive bullet points preserving exact figures and decisions. Use when producing a human-readable executive summary of a single provided document; for researching a topic across the live web with Google Search citations, use synthesize_web_intelligence_brief instead. Stateless read-only edge call (max 32 KB, 60 RPM): accepts 'text' ('summary_format' defaults to 'executive_bullets'), supports '?demo=true' or L402 (100 sats), and returns HTTP 400 if invalid.

Synthesizes a real-time market or technical intelligence brief grounded in live Google Search, returning strict JSON ('executive_summary', 'key_developments', 'sentiment', 'sources'). Use when fetching a structured JSON market/news brief with live web citations and sentiment classification; for summarizing caller-supplied text offline without live web search, use summarize_technical_document instead. Read-only edge call (32 KB, 60 RPM): accepts 'topic' ('focus' defaults to 'technical_and_market_catalysts'), supports '?demo=true' or L402 (200 sats), and returns HTTP 400 if invalid.

Audits a compacted conversation summary against the original full context to detect dropped user constraints, lost file paths, altered numbers, or missing pending tasks. Use when verifying that an LLM context-compaction summary retained all task-critical facts before truncating history; for generating a new executive summary from a raw technical document, use summarize_technical_document instead. Read-only edge call (32 KB, 60 RPM): accepts 'original_context' ('compacted_summary' defaults to 'standard'), supports '?demo=true' or L402 (90 sats), and returns HTTP 400 if invalid.

Maps a natural-language agent goal to an ordered sequence of Diogenes MCP tool calls with exact endpoint paths, JSON-RPC 2.0 payloads, and cumulative satoshi cost estimates. Use when planning which MCP tools to chain together to solve a multi-step objective within a satoshi budget; for compiling a static XML system prompt rather than an executable MCP tool call sequence, use compile_xml_agent_prompt instead. Read-only edge call (32 KB, 60 RPM): accepts 'objective' ('budget_preference' defaults to 'lowest_sats_deterministic'), uses '?demo=true' or L402 (100 sats); HTTP 400 if invalid.

Evaluates a worker agent's submitted task deliverable against agreed contract acceptance criteria to issue a deterministic escrow settlement verdict ('RELEASE_FUNDS_TO_WORKER_AGENT', 'REQUEST_REVISION', or 'REFUND_CLIENT'). Use when settling multi-agent bounties or verifying subcontractor deliverables before releasing MPP/L402 escrow funds; for checking syntactic JSON Schema property types without evaluating semantic contract terms, use enforce_json_schema_types instead. Read-only call (32 KB, 60 RPM): accepts 'deliverable' (with enum defaults), uses '?demo=true' or MPP (150 sats); HTTP 400 if invalid.

Initializes a Base Native ETH Machine Payments Protocol (MPP) metered inference session, allocating a streaming token quota and returning a step-by-step execution plan. Use when running session-metered agent planning settled via Base EVM MPP 'Payment-Challenge' headers; for mapping a task goal to L402 MCP tool calls without opening an MPP streaming session, use route_mcp_workflow_plan instead. Stateless read-only edge call (max 32 KB, 60 RPM): accepts 'task' ('token_quota_tier' defaults to 'compact_1024'), supports '?demo=true' or MPP (500 sats), and returns HTTP 400 if invalid.

Generates normalized high-dimensional semantic vector embeddings settled via Solana SPL USDC ('x402' protocol headers) for decentralized RAG pipelines. Use when computing text embeddings paid via Solana Native SOL x402 signatures rather than Bitcoin Lightning; for extracting discrete JSON entities from unstructured text, use extract_structured_json_entities instead. Stateless read-only edge call (max 32 KB, 60 RPM): accepts 'text' ('encoding_format' defaults to 'float_array'), supports '?demo=true' or x402 (100 sats), and returns HTTP 400 if invalid.