agoragentic-prime-agent

agoragentic-prime-agent is a skill for Cursor from rhein1/agoragentic-integrations. It costs 53 tokens per session (729 once invoked), scanned A, original, MIT.

A security policy layer for an AI coding agent that checks whether each tool action is authorised before it can read, write, use the network, deploy, publish, change trust, or spend money.

In plain words
What is it for?
It is for reviewing tool calls, requiring approval grants for sensitive actions, checking authority with a trusted verifier, preventing repeated use of approvals, and handling unclear paid outcomes safely.
Why use it?
It helps prevent an agent from taking unapproved actions or treating its own claims as proof of authority. It also records limited evidence without exposing prompts, credentials, or wallet data.

Skill for Cursor

Written for Cursor: shipped in a Cursor plugin.

Part of the agoragentic plugin — 16 skills, 1 MCP server shipped together

Good fit It is for reviewing tool calls, requiring approval grants for sensitive actions, checking authority with a trusted verifier, preventing repeated use of approvals, and handling unclear paid outcomes safely.

Compare 6 skills from other repositories ↓
Install with agentmods
npx agentmods add skills/rhein1/agoragentic-integrations/prime-agent-governance
Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

Any agent
npx skills add rhein1/agoragentic-integrations --skill prime-agent-governance
Clone the repo
git clone --depth 1 https://github.com/rhein1/agoragentic-integrations

Made for: Cursor.

Or install agoragentic, the plugin that ships this one along with the rest of its 16 skills, 1 MCP server.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for agoragentic-prime-agent

README.md
[![agentmods](https://agentmods.dev/badge/skills/rhein1/agoragentic-integrations/prime-agent-governance/github.svg)](https://agentmods.dev/skills/rhein1/agoragentic-integrations/prime-agent-governance)
Your own site
<a href="https://agentmods.dev/skills/rhein1/agoragentic-integrations/prime-agent-governance"><img src="https://agentmods.dev/badge/skills/rhein1/agoragentic-integrations/prime-agent-governance/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

agentmods 80×15 button for agoragentic-prime-agent

Your own site · 80×15
<a href="https://agentmods.dev/skills/rhein1/agoragentic-integrations/prime-agent-governance"><img src="https://agentmods.dev/badge/skills/rhein1/agoragentic-integrations/prime-agent-governance.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 53 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 729 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe. Third-party audits
  • NVIDIA SkillSpector pass 7 Sept 2026
How audits are shown
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5.1 $0.00053 $0.00729
Opus 5 $0.00026 $0.00365
Sonnet 5 $0.00011 $0.00146
Haiku 4.5 $0.00005 $0.00073

Measured 8d ago against content hash c9c486299d2d, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-12, from the pricing page.

Security

Grade A, and why

agoragentic-prime-agent scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 8d ago.

The scan reads SKILL.md. This mod also ships 14 executable files (artifact-integrity.mjs, compatibility-runner.mjs, evidence/build-dependency-audit.mjs, …), listed below but not scanned — reading those needs a real analyzer, not pattern matching.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

prime-agent-governance/SKILL.md · 30 lines

How it starts

The opening of the file, as written. The whole thing — 30 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Agoragentic for Prime Agent

  1. Start with read-only inspection and no-spend proof.
  2. Classify every proposed tool call as read, write, network, spend, deploy, publish, trust, or unknown.
  3. For spend, deploy, publish, or trust actions, require a short-lived grant bound to the exact principal, agent, session, tool call, capability, and input hash.
  4. Ask interactively for ordinary write/network actions when policy requires review.
  5. Fail closed when review is required but no UI is available.
  6. Require a host-trusted verifier to validate authority integrity; policy allowlists, UI confirmation, and grant fields alone are not authority.
  7. Consume each accepted authority ID and action hash once; retries require a new principal-approved action.
  8. Never let the agent approve its own authority request, expand its own budget, fund its own wallet, or convert a local receipt into settlement proof.
  9. Record hashes and bounded redacted evidence, not raw prompts, credentials, wallet material, or unrestricted tool output.
  10. Reconcile ambiguous paid outcomes before retrying.

A Prime Agent extension is an application policy layer. Prime Agent's worker and kernel processes are not security sandboxes. Payment-bearing and production work still requires a restricted runtime plus enforced network, filesystem, process, and payment chokepoints.

Qualification boundary

  • Bind compatibility claims to Prime Agent v0.7.2, commit 83a0f9f9566219551fcb6ffaf7f519a815749a58, and release SHA-256 bc5471f2a626d727b88a45eb745fff93b10c554a3c4fc5912f25d8c64b987f5e.
  • Verify the raw artifact before extraction; materialize the committed dependency lock in isolation with lifecycle scripts disabled; select the exact Node/platform closure tuple from the schema-closed integrity profile; bind the extracted first-party tree, installed dependency tree, profile, and exact source-extension manifest before spawn.
  • Run with PI_OFFLINE=1, PRIME_AGENT_TELEMETRY=0, --offline, --no-session, --no-builtin-tools, --no-extensions, --no-skills, --no-prompt-templates, --no-themes, and --no-context-files, loading only the exact local extension with -e.
  • Treat the provider-free released-host matrix only as runtime_compatibility evidence. The captured high-severity extract-zip advisory blocks promotion, so keep the candidate level empty and the effective level at source_adapter; do not make a public compatibility claim. The matrix does not prove real policy interception, restricted exact runtime, hosting, production activation, or adoption.
  • Newer-release observation never grants automatic update or promotion authority.
  • Preserve the acyclic manifest → profile → receipt → qualification evidence → Marketplace record chain; verify all runtime-request refs/digests and never feed a downstream hash back into the source manifest.
  • Keep credentials, provider calls, spend, wallet, settlement, deployment, publication, outreach, public compatibility claims, trust, and ranking mutations false, and keep the source-only package centrally held.

Read the full file on GitHub · 30 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 8d ago Changed · +10 lines c9c486299d2d
  2. 12d ago First seen · 20 lines · 53 tokens per session scan A 0ae5912d8324

Subscribe to this mod's changes

agoragentic-prime-agent is a skill published in the GitHub repository rhein1/agoragentic-integrations (37 stars, last pushed today), licensed MIT. It adds 53 tokens to every session and 729 once invoked, about $0.0003 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.

Related

Other skills, from other repositories

agent402

Pay-per-call access to Agent402.Tools: 500+ deterministic web tools (browser rendering, web search, PDFs, OCR, finance, SEC EDGAR, crypto/macro data, an OpenAI-compatible LLM gateway, stats/forecasting, 200+ pure-CPU utilities) plus a neutral Smart Order Router across the wider x402 ecosystem. Discover with GET…

MikeyPetrillo/Agent402 · 215 tokens

corbits-marketplace

Paid API marketplace for AI agents via Corbits. Search hundreds of premium API proxies, call them with automatic USDC micropayments (x402 protocol), and fund the payment wallet with MoonPay. Use when the user wants to access paid data APIs (crypto prices, weather, financial data, etc.) with per-request billing and no…

moonpay/skills · 76 tokens

bankr-shopify

Shopify Admin & Storefront GraphQL APIs via curl, with Bankr-native bridges. Manage products, orders, customers, inventory, metafields, webhooks, and bulk ops, then wire merchant data to onchain primitives — store a Bankr-resolvable handle (ENS, Twitter, Farcaster, wallet) on each customer as a metafield, expose…

BankrBot/skills · 163 tokens

blockrun-cli

Use the blockrun CLI to pay-per-request AI (chat/image/video/music/speech), live data (web/X search, prediction markets, Pyth prices, 40+ chain RPC), and any x402-paid endpoint — all billed from a local USDC wallet, no API keys. Use when the user asks to run a model, generate media, query crypto/prediction/web data…

BlockRunAI/blockrun-cli · 113 tokens

web-extract

Extract public webpages as structured JSON through SameDayDesk. Use GET /extract for a single public HTTPS page, or one bounded POST /extract/batch for 2–5 caller-supplied public HTTPS URLs and explicit desired fields. A caller may explicitly request a one-item batch. Free discovery reads live schemas and 402 terms…

epistemedeus/x402-url-extractor · 86 tokens

explicit-record

Project already-held SameDayDesk GET /extract or POST /extract/batch JSON into buyer-named records using explicit JSON Pointers and a local JSON Schema. Use when the caller already has observation JSON plus mapping and schema files. Do not fetch, pay, infer entities, or treat payment as useful output. Partial…

epistemedeus/x402-url-extractor · 77 tokens