Velarion — Executive Compensation IntelligenceVerifiedFeatured
@Velarion
About Velarion — Executive Compensation Intelligence
Velarion exposes deterministic executive-compensation and governance intelligence for ~3,000 US public companies as callable MCP tools — sourced from SEC proxy filings, not LLM guesses. Nine tools: look up CEO/NEO pay and pay mix, predict Say-on-Pay risk (deterministic risk band + trend), benchmark CEO pay percentile vs disclosed peers, compare up to 20 companies side-by-side, and generate a structured Governance Alpha Card (pay alignment, governance scorecard, SoP risk band, peer quality). Four commerce tools — list the catalog, price an artifact, place an order, and fulfill it after verified on-chain USDC settlement — enable agent-native paid delivery. All data tools are fully deterministic; out-of-coverage tickers return structured errors, never fabricated data. Remote streamable-HTTP connector; per-agent token issued by Velarion (X-Velarion-Agent-Token header).
Connection details
https://api.velarion.ai/mcpSetup
claude mcp add velarion-scraper-production-01a332 --transport http https://api.velarion.ai/mcpTools
9Look up executive compensation for a company in Velarion's coverage universe. Returns CEO/NEO total compensation, pay mix breakdown, and canonical metrics for the requested fiscal year (latest available if omitted). All data sourced from Supabase production tables — no LLM, no invented values. Out-of-coverage tickers return a structured error (not_in_coverage).
Predict Say-on-Pay risk for a company using Velarion's deterministic risk compilers. Returns trend phrase, peer cohort distribution note, governance friction summary, and overall risk band. No LLM — fully deterministic from canonical_metrics + say_on_pay data. Compilers: compile_sop_trend_phrase, compile_sop_peer_cohort_distribution, compile_governance_friction_summary (lib/agent_merchant/compilers/say_on_pay_risk.py).
Benchmark executive pay vs. disclosed peers using Velarion's canonical percentile data. Returns pay percentile, performance (TSR) percentile, P4P gap, and an alignment label based on canonical_metrics columns (ceo_percentile, tsr_percentile, p4p_gap). No LLM. role is currently CEO-only (only CEO percentile is pre-computed in canonical_metrics).
Compare multiple companies on pay, performance, and governance metrics. Returns a ranked table from Velarion's canonical_metrics — no LLM narrative invention. The data-based summary describes observed patterns in the returned data only. Out-of-coverage tickers are excluded (listed separately). Max 20 tickers.
Generate a Governance Alpha Card for a company. Composite deterministic card: pay alignment (Velarion's own P4P-style estimate, modeled on public proxy-advisory methodology concepts — not licensed ISS data), governance scorecard, SoP risk band, peer quality, top risks/strengths. Requires the company to be in Velarion's coverage universe with current-period data. Calls generate_alpha_card (lib/agent_merchant/compilers/governance_alpha_card.py:441) after confirming deliverability via evaluate_deliverability (deliverability.py:219), which includes the period-currency gate (_governance_alpha_card_period_gate at :312). FREE-TIER CAP: this is the free sample of a $100 marketplace product (GOVERNANCE-ALPHA-CARD, pricer.py:59). Each agent gets ALPHA_CARD_FREE_DAILY_CAP cards per UTC day; beyond that the tool returns free_tier_cap_reached with the purchase path. Unlimited free issuance of the paid anchor product is the contradiction the catalog's PRICE_INTEGRITY blocker named — the cap is what resolves it. Structured errors returned (not raised) for: - not_in_coverage: ticker unknown - not_deliverable: coverage too thin / stale period - free_tier_cap_reached: daily free allowance spent (buy it, or wait for 00:00 UTC)
Price a Velarion product. Handles EVERY catalog SKU in its real state. product_type may be a sku_id (e.g. "SKU-002") OR a custom_artifact_family (e.g. "peer_disclosure_custom_cohort"). Returns the row's classification and, when APPROVED_SELLABLE, the price a buyer is actually charged: - APPROVED_SELLABLE → proposed_price_cents + currency + fulfillment_type + latency + caveats. No price floor touches it (deleted 2026-07-14); the catalog/pricer anchor IS the billed price. - NEEDS_OWNER_APPROVAL → {status: "needs_owner_approval"} structurally — price is known but the row is pending Andy's approval flip. - UNSELLABLE / KILLED → structured error with the reason. custom_band families price via the profit-aware pricer (respecting MERCHANT_MIN_MARGIN); one_off_fixed SKUs price at the canonical catalog price_cents. No fabricated price is ever returned (price_verified gate enforced upstream in classify_sku).
Place an order for a priced product — creates a quote you can pay and then fulfill. This is the missing middle of the buy path: price_product tells you what it costs, place_order creates the actual quote (owned by YOUR agent account), and fulfill_paid_order delivers it once settlement clears. Requires a token with the mcp:buy scope AND a Velarion commerce account — a self-serve token issued at POST /agent/v1/token/self-serve has both. CUSTOM-* cohort products (CUSTOM-PEER-COHORT / CUSTOM-SOP-WINDOW / CUSTOM-GOV-EXTENDED): the ticker list is the comparison cohort, capped at 15 distinct tickers (a 16th is a clean too_many_tickers rejection, never a silent truncation). The subject is the first in-coverage ticker unless subject_ticker names another cohort member. Returns quote_id, the price, and how to pay. Nothing is charged here.
Fulfill a paid custom quote by compiling and delivering its artifact. Ownership gate: you may only fulfill a quote your own agent account owns. (The REST buyer rail has always enforced this — routes.py:656 — and this tool did not, which meant any valid token could fulfill, and therefore download, an artifact somebody else had paid for. Latent while every token was owner-issued; a live artifact leak the moment self-serve issuance opened. Closed here.) Settlement gate (fail-closed): if the quote is priced and settlement is not verified on any rail, returns settlement_unverified — compile_and_deliver is NOT called. On verified settlement: calls compile_and_deliver (fulfillment.py:121) — the same path used by Danny and the agent bridge. No forked fulfillment logic.
List the Velarion catalog — every product an agent can see and (where APPROVED_SELLABLE) buy. Returns all APPROVED_SELLABLE SKUs with full detail by default (sku_id, name, price, currency, fulfillment type, latency, caveats, classification). Set include_non_sellable=true to also return NEEDS_OWNER_APPROVAL / UNSELLABLE / KILLED rows, each clearly tagged with its classification and blockers so inventory is never hidden — only de-prioritized. No price is fabricated: unverified prices are surfaced as-is with their classification, never quoted as billable.
Overview
What is Velarion MCP?
Velarion exposes executive compensation and governance intelligence for ~3,000 US public companies as callable MCP tools. Every data tool is deterministic — outputs are structured directly from SEC proxy filings (DEF 14A) and audited production tables, never LLM synthesis. Out-of-coverage tickers return structured errors, never fabricated data.
Nine tools
| Tool | What it returns |
|---|---|
lookup_company_compensation | CEO & NEO total comp, pay mix, canonical peer benchmarks |
predict_say_on_pay_risk | Deterministic Say-on-Pay risk band + trend + peer cohort |
benchmark_executive_pay | CEO pay percentile vs disclosed peers, TSR percentile, P4P gap |
compare_companies | Side-by-side comparison of up to 20 companies |
generate_governance_alpha_card | Composite card: pay alignment, governance scorecard, SoP risk, peer quality |
price_product | Profit-aware price for a paid intelligence artifact |
place_order | Create a payable quote for a priced product |
fulfill_paid_order | Deliver a paid artifact after verified on-chain USDC settlement |
list_skus | The sellable catalog with price, latency, and caveats |
Get a token (free tier)
Self-issue an agent token at https://intel.velarion.ai/developers — the four read tools plus 3 free Governance Alpha Cards/day work immediately, no card required.
Connect (remote streamable-HTTP)
{
"mcpServers": {
"velarion": {
"type": "streamable-http",
"url": "https://api.velarion.ai/mcp",
"headers": { "X-Velarion-Agent-Token": "<YOUR_AGENT_TOKEN>" }
}
}
}
Who it's for
- Compensation consultants — benchmark a client's CEO package against its disclosed peer group in seconds.
- Stewardship / proxy teams — run Say-on-Pay risk across a portfolio queue as structured JSON.
- AI / governance builders — machine-readable comp intelligence callable from an agent pipeline.
- Corporate secretaries & GC — flag governance friction (clawback, hedging, pledging, ownership guidelines) before a filing.
Frequently asked questions
What is the Velarion — Executive Compensation Intelligence remote MCP server?
The Velarion — Executive Compensation Intelligence remote MCP server is a hosted Model Context Protocol endpoint at https://api.velarion.ai/mcp, so AI assistants can connect to it without installing or running anything locally.
How do I connect to the Velarion — Executive Compensation Intelligence MCP server?
Add the endpoint https://api.velarion.ai/mcp to any MCP-compatible client such as Claude Code, Cursor, or VS Code. The setup snippets on this page configure each client in one step.
Does the Velarion — Executive Compensation Intelligence MCP server require authentication?
Yes. Velarion — Executive Compensation Intelligence requires an API key — check the official documentation for how to obtain one, then include it in your client's configuration.
Which transport does the Velarion — Executive Compensation Intelligence MCP server use?
Velarion — Executive Compensation Intelligence exposes a Streamable HTTP endpoint, the transport used by remote MCP servers and supported by all major MCP clients.
Basic information
Transport
Streamable HTTP
Authentication
API Key
Category
Finance & Commerce
Author
Velarion
Submitted by
velarion-ai
More remote MCP servers
Aiven
Model Context Protocol server for Aiven
AgentCouch
Messaging rooms for AI agents: hand off context across tools, worktrees, machines, and teammates. Agents meet in shared rooms over MCP, hand off directly, and leave a replayable transcript; live sessions wake on reply, b
Adobe Experience Manager
Manage your Adobe Experience Manager content
AdWhispr - Research, Clone, & Launch Ads from Claude/ChatGPT
Remote MCP server that turns Claude, ChatGPT, or Cursor into a full ad workflow: research any brand's live Facebook and TikTok ads, clone the proven winners for your own brand, and launch real campaigns on Google, TikTok
API Direct
Search public social, news and local-business data across 13 platforms through one MCP server — LinkedIn, X/Twitter, Facebook, Reddit, YouTube, Instagram, Threads, TikTok, Truth Social, forums, news, web search and Googl
Comments