All projects on Base matching “execution”

DCL Trust Oracle. Call only when the text must include the substring AI and must not include "I cannot be held responsible" or "no guarantees". POST JSON {"response","agent_id"}. COMMIT means those checks passed and the action may proceed. NO_COMMIT means do not execute the checked action. Not a tool-call guardrail, prompt-injection detector, or jailbreak screen; use /evaluate/fast for the phrase

Decide whether an agent-authored PostgreSQL statement complies with an execution policy, using PostgreSQL's own parser rather than pattern matching. Detects statement stacking, data-modifying CTEs, COPY PROGRAM, privilege changes, and dangerous functions that read as ordinary SELECTs. Returns reason codes and an optional short-lived signed permit bound to the exact normalized statement. Never conn

Latest releases of the 10 main Ethereum clients as JSON: execution (go-ethereum, Nethermind, Besu, Erigon, Reth) and consensus (Lighthouse, Prysm, Teku, Nimbus, Lodestar). Version, date, days since release, link to notes, and a plain keyword flag when notes mention security or urgent upgrades.

Check whether an approved Base token swap is eligible before supplying a wallet. Returns verified token contracts, indicative output, liquidity sources, fee estimate, warnings, and the exact next action. No custody, signing, or execution.

Darrylbots paid research: Standard Stock Screen: Quality, Momentum, and Durable Execution

Datoka Agent: one signed execution declaration receipt. Retry with the same event, idempotency key and payment.

Decide whether an agent-authored bash command complies with an execution policy, using a real shell tokenizer rather than pattern matching. Catches recursive deletes, force pushes, cluster and infrastructure teardown, raw device writes, path escapes, privilege escalation, and curl-piped-to-shell through pipelines, substitutions, quoting games and base64. Returns reason codes with spans and an opti

VpsAgent — rent disposable VPS compute via x402, no signup required. Two product lines: ephemeral sandboxes (1h / 6h / 24h, 1 vCPU / 1 GB RAM / 25 GB SSD, $0.50–$5.00) for code execution, CI runs, browser automation; persistent servers (small / medium / large, 30-day or 365-day, $14.99–$449.99) with 4–8 vCPU / 8–24 GB RAM / 100–300 GB SSD, 200–600 Mbit/s unmetered for hosting agents, services, dev

RAVN — execute a previously quoted swap. Returns a signable transaction, typed data to sign, or a Bitcoin deposit address depending on the route; RAVN never custodies funds. Same 0% fee, non-custodial engine the RAVN app itself runs on.

Execute code in any supported language. 5s CPU timeout, 128MB RAM. Supports 40+ languages including Python, JavaScript, C, C++, Java, Go, Rust.

Everything about a Base transaction in one call: status, time, from and to, ETH value, the function called (for common selectors), the fee split into L2 execution and L1 data in ETH and USD, every ERC-20 transfer with symbol and readable amount, NFT transfers, and x402 payments (USDC transferWithAuthorization) with the seller named from Bazaar.

AI-agent execution verification and automated task acceptance: evaluate caller-supplied functional, security, privacy, compliance, transaction, abuse, regression and economics gates; ARN does not perform the underlying checks. Re-run for each release or evidence change; sustained CI users can switch to the 30-day Release Guard pass.

Compute a SHA-256 digest for customer-supplied UTF-8 or base64-decoded bytes through a stable machine endpoint. Returns the reproducible digest through a receipted external execution.

One crypto allocation per day, fixed after the daily close. Since 12 September 2026 each day is fingerprinted on Sui before execution; earlier days were fingerprinted together on 11 September. Past days per call; today's allocation from 07:00 Zurich per call or with a pass. Each day from 2 October 2026 can be checked against Sui without trusting the seller.

Interzoid Batch Data Enrichment API: enrich up to 100 JSON records per call with AI company intelligence (profiles, parent companies, executives, competitors, buying signals, funding, ESG, gov contracts, tech stacks, NAICS/SIC codes), email, phone, and IP scoring, sales tax, VAT, and customs rates, property data, standardization, translation, or custom fields. 39 functions, 16 languages, processed

Extract a readable article from supplied HTML, including title, author, language, metadata, clean Markdown, links and tables. Uses Mozilla Readability with a fallback for short pages. No remote fetches or script execution; URLs resolve relative links.

Exit-liquidity and price-trust assessment for a Base ERC-20 token at a requested USD position size. On-chain sell quotes cover Aerodrome, Slipstream, and Uniswap V2/V3/V4. Returns tradeable, caution, illiquid, or unverified with executable proceeds, price impact, confidence, concerns, and unmeasured checks. Inputs: token address and size_usd.

Before an AI coding agent (or you) runs a GitHub repo's install steps: scans README, package.json install scripts, install.sh/setup.py, .vscode/tasks.json and Makefile for download-and-execute one-liners (iwr|iex, curl|sh), obfuscated or base64 payloads, known malware domains, typosquatted crypto dependencies, agent-targeted prompt injection and password-protected archives, plus GitHub signals of

Problem: Select a supported execution plan against the supplied cost, depth, duration limits, and weights. Input: JSON with plans, weights, maximum cost usd, maximum depth, maximum duration seconds. Result: ranked feasible plans, objective scores, selected plan identity. Limits: Local software evidence only; 4096 operations; 65536 request bytes.

Prediction-market divergence feed: the same event priced differently on Kalshi vs Polymarket. Returns matched events whose last prices differ by at least min_gap_pp points, with both venues bid/ask, the last-price gap and the executable gap (rich venue bid minus cheap venue ask), and a match-confidence score. The benchmark is the other venue price for the same event; matches are heuristic, so chec

Deterministic structural failure inspection of a caller-declared control mesh. Structural evidence only; never authorizes execution. 0.01 USDC per request.

Native gas on Ethereum, Arbitrum, Base, BNB Chain, Polygon or Optimism, paid in USDC on Base, Arbitrum or Polygon. Offers come from /x402/gas/quote for the chains, a recipient and a USDC amount. The payment goes to an address made for that one order, which delivers at least the stated minimum in native; from another chain, USDC on arrival if the arrival swap fails. An order not executed by its dea

Instant Cert — LEGACY TASTE ONLY Exact $0.05 for ONE mutating SQL PASS/FAIL. OFF TABLE as wealth. Prefer Authorize mutating SQL before execute via Pilot Challenge unlock POST /v1/challenge/unlock $100 (amount 100000000) or Gateway invoice $299/mo. Not a production authorize path.

Pre-payment counterparty risk check for AI agents: OFAC sanctions (names, aliases, wallets), domain age and email authentication, vendor and executive impersonation domains, routing/IBAN/SWIFT/wallet validation, and invoice-fraud signals. Returns proceed/review/block, explained findings and a signed receipt.

Execute a structured FORGE task using a promoted tool

General current-operation routing for permit/exemption/duty-of-care and Digital Waste Tracking review using the same deterministic engine as the MCP tool. Payment verification precedes deterministic execution. Not Environment Agency approval or legal advice.

Paid single sponsor-change impact preflight. Deterministic Skilled Worker sponsor-change impact preflight with official-source freshness gates. Payment verification precedes business execution. Not legal advice or a Home Office decision.

Buy a live status line install for a team. Returns the client script file(s), config snippets, and step-by-step instructions for the installing agent to execute locally. The response token makes the line endpoint work and re-downloading the same bundle via /api/bundle is free — never pay twice for a retry.

Statically analyse a Model Context Protocol (MCP) server manifest for security risks. Tool descriptions enter the model context, a prime prompt-injection vector. Deterministic and offline (inline manifest). Flags dangerous capabilities (command/code execution, filesystem, credential and network access), injection patterns, hidden unicode and unconstrained parameters. Returns severity-ranked findin

Static malware and prompt-injection scan of an OpenClaw / AgentSkills bundle before install. Detects credential and wallet theft, exfiltration endpoints, remote code execution, reverse shells, obfuscated payloads, persistence, cryptominers, and injection in SKILL.md. Returns PASS/REVIEW/BLOCK with risk score and per-finding file, line and evidence. bundle = {files:[{path,content}]} or {github:'own