- Nebelus Agent Construction @nebelus/constructionBuild & govern AI agents for regulated industries: EU/GCC residency, governance, audit.7% chance this is malicious. Assessed by AI, which can make mistakes.Runs on your machine, installed from a package registry.
- SALT19 Agent Utility Grid @gabrielallit1/salt19-agent-utility-gridMCP/x402 tools with bounded decisions, adaptive pricing, and Proof402 machine-commerce receipts.9% chance this is malicious. Assessed by AI, which can make mistakes.A remote server; no sign-in.
- ARBITER @grip/arbiterDecision arbitration, action routing, workflow control, deployment remediation, persistent memory.10% chance this is malicious. Assessed by AI, which can make mistakes.A remote server; no sign-in.
- Handoff — shareable agent result pages + human-in-the-loop approval gates @notdown-app/human-approval-decision-gate-and-shareable-result-page-publisherFollowable agent result pages + human-in-the-loop yes/no approval gates before irreversible actions.6% chance this is malicious. Assessed by AI, which can make mistakes.Runs on your machine, installed from a package registry.
- PHION Agent Trust Infrastructure @phion/evidence-engine123 agent services with free discovery, intent, x402 terms and signed verification.8% chance this is malicious. Assessed by AI, which can make mistakes.A remote server; no sign-in.
- Autonomous Agents Network (AANet) @aanet/aanet-mcpMulti-agent coordination outside your LLM context: dedupe parallel work, per-agent budgets, handoff6% chance this is malicious. Assessed by AI, which can make mistakes.A remote server; no sign-in.
- ActionDock @actiondock/actiondockSupervised API-write gateway for AI agents with policy, human approval and execution receipts.5% chance this is malicious. Assessed by AI, which can make mistakes.A remote server; no sign-in.
- Spendline @spendline/spendlineEnforce AI budgets before the model call and track cost per customer across 10 providers.4% chance this is malicious. Assessed by AI, which can make mistakes.A remote server; no sign-in.
- FLOCORE Governed MCP @flocore/flocoreGoverned MCP front door to FLOCORE; write calls need human approval and signed receipts.5% chance this is malicious. Assessed by AI, which can make mistakes.A remote server; no sign-in.
- S+S Agentic @shortandsweet/doorsDoors for AI agents: witness, letter, poison check, ghost check, wall. No account, no payment.8% chance this is malicious. Assessed by AI, which can make mistakes.A remote server; no sign-in.
- Cromanion @cromanion/cromanionRaise conversion on a site you manage. No API key, no human to wake — you prove the domain.15% chance this is malicious. Assessed by AI, which can make mistakes.A remote server; no sign-in.
- Vigil — an outside witness for AI agents @vigilnotary/vigilAgent leases, signed receipts, watchdogs, and optional paid source-verification evidence.7% chance this is malicious. Assessed by AI, which can make mistakes.A remote server; no sign-in.
- Hunter-Seeker @hunter-seeker-io/hunter-seekerThe decision layer for AI agents: act, escalate or refuse, and every decision comes back signed.11% chance this is malicious. Assessed by AI, which can make mistakes.A remote server; you sign in with your account.
- 402Signal @402signalhq/402signalCheck paid API offers before your agent spends. Qualifying routes: $0.003 USDC. Normal misses: free.9% chance this is malicious. Assessed by AI, which can make mistakes.A remote server; no sign-in.
- Traced Claude Format Guard @getvda/traced-claude-format-guardTraced Claude Format Guard13% chance this is malicious. Assessed by AI, which can make mistakes.A remote server; no sign-in.
- Vizier @vassiliylakhonin/vizierDeterministic authorization for one proposed AI agent action, returned with a signed receipt.4% chance this is malicious. Assessed by AI, which can make mistakes.A remote server; no sign-in.
- Batcave Detective Basic @bjorn2-sol/batcave-detectiveEvidence-bound second-opinion audit of an agent conclusion against caller-supplied evidence.11% chance this is malicious. Assessed by AI, which can make mistakes.A remote server; no sign-in.
- My Fordyce @fordycecg/my-fordyceThe backlog that hands out the work: multi-agent task coordination — leases, claims, gates.9% chance this is malicious. Assessed by AI, which can make mistakes.A remote server; you sign in with your account.
- dibs @polymatx/dibsCoordination for parallel coding agents: file claims, enforcement hooks, git-native lessons.6% chance this is malicious. Assessed by AI, which can make mistakes.Runs on your machine, installed from Docker.
- Preflight SafeSwap @mr8-gif/preflight-safe-swapSafe Solana swaps for AI agents with risk checks, unsigned transactions, x402, and 3-bp fees.6% chance this is malicious. Assessed by AI, which can make mistakes.A remote server; no sign-in.
- HumanMirror One @humanmirror/oneAgent routing, safety preflight, execution receipts, best execution and verified outcomes.13% chance this is malicious. Assessed by AI, which can make mistakes.A remote server; no sign-in.
- Sentarion @sammyboi81/sentarionGoverned MCP orchestration, tamper-evident memory, and Claude Code Seatbelt hooks.7% chance this is malicious. Assessed by AI, which can make mistakes.Runs on your machine, installed from PyPI.
- Toreva Goals @toreva/goalsDurable, user-controlled goals and governed plans for AI agents.7% chance this is malicious. Assessed by AI, which can make mistakes.A remote server; you sign in with your account.
- BrandPresence @nuteee/brandpresenceOrganization-scoped BrandPresence evidence and explicitly authorized product actions over MCP.10% chance this is malicious. Assessed by AI, which can make mistakes.A remote server; you sign in with your account.
