Loading…
Tradeable signals from public-company website stack changes for quant research: 7 tools, no auth required.
Do not connect
A critical issue was found. Do not connect this server as-is.
Scanned 1 month ago Due for re-check
A server can change after it's graded. Re-run the automated scan to refresh this report.
This grade is deterministic and reproducible: the same server surface always yields the same grade under a given algorithm version. It is a real automated assessment computed by the MCPGrade engine from what the probe actually observed — not a fabricated or opinion score. It is not a manual human pentest, so it can miss context-specific risks.
Every signal below was measured directly by the automated probe. The grade is derived only from evidence like this — nothing is assumed.
The boolsai-signals MCP server exposes 12 tools, focused primarily on general-purpose capabilities. Its published description reads: "Tradeable signals from public-company website stack changes for quant research: 7 tools, no auth required". It communicates over Streamable HTTP using the 2026-07-28 protocol revision, and does not require authorization to connect. MCPGrade currently rates boolsai-signals F — a critical issue was found and the server should not be connected as-is. Its most notable findings include "Cross-tool shadowing" and "Data-exfiltration parameters". This report is a deterministic, reproducible automated assessment: the same observed surface always yields the same grade under a given algorithm version, and it is refreshed as new scans arrive and free to read — but it is not a substitute for a manual human security review. Always re-check boolsai-signals's advertised tools, transport security, and authorization posture before connecting an autonomous agent or sharing sensitive context with it.
Add boolsai-signals to an MCP client using the endpoint below. Review the grade and findings above before granting an autonomous agent access.
MCP endpoint
https://signals.boolsai.ai/mcpStreamable HTTP transport. This is the MCP endpoint, not a website — paste it into your MCP client server list rather than a browser.
Captured passively during the read-only scan. Click any value to find servers that match it.
The tools this server advertises via tools/list — names, purposes, and the parameters each accepts, exactly as enumerated read-only.
Orient the agent: total events, tickers, date range, top event types, top detectors, price coverage, SPY benchmark status. Call this FIRST when starting research. Returns counts that let the agent rea
Automated pattern discovery — scans event_type × detector × diff_field × severity combinations and returns those with the strongest forward-return characteristics (α vs SPY, % positive, n). Use this w
Compute α stats for an arbitrary filter expression. Use this to test a specific hypothesis (e.g. 'tier_count_changed on enterprise-SaaS tickers' or 'severity 5 events that happened on Mondays'). Retur
Sorted worst-first. Each finding shows its severity, what it means, its OWASP MCP Top-10 mapping, and a recommended fix — the check id links to the exact methodology row that produced it.
tool: event_dossier
A tool description tries to alter the model’s use of another tool.
vent flagged by find_signals or recent_events.RecommendationDescriptions must describe only their own tool.
tool: scan_at_date
A tool exposes a free-text/webhook/feedback parameter with no honest purpose — a classic exfil channel.
parameter "url"RecommendationRemove side-channel parameters; constrain tool inputs.
tool: wayback_backtest
A tool description tries to alter the model’s use of another tool.
nstead of the recent live event dataset (2 months, 1.7K events). Much biggRecommendationDescriptions must describe only their own tool.
tool: signal_landscape
A tool description tries to alter the model’s use of another tool.
having to call find_signals N times. Stateless, pure D1, no rate-limit riskRecommendationDescriptions must describe only their own tool.
tool: farm_domain
A tool description tries to alter the model’s use of another tool.
dy farmed (i.e. wayback_backtest / domain_timeline return no data for it). HitsRecommendationDescriptions must describe only their own tool.
tool: (server instructions)
A tool exposes a free-text/webhook/feedback parameter with no honest purpose — a classic exfil channel.
** — 2 years of history) — weekly diffs from wayback scans. WAY biggerRecommendationRemove side-channel parameters; constrain tool inputs.
tool: (server instructions)
A tool description tries to alter the model’s use of another tool.
fficiency) 1. `universe_summary` — orient (counts, date ranges, top tickers/eveRecommendationDescriptions must describe only their own tool.
The server accepts tool enumeration (and likely invocation) with no authentication.
RecommendationRequire OAuth 2.1 authorization for any server exposing non-public tools.
tool: find_signals
The server advertises open-world / broadly-scoped capabilities.
no annotations on write-capable tool "find_signals"RecommendationScope tools to the minimum needed.
tool: signal_landscape
The server advertises open-world / broadly-scoped capabilities.
no annotations on write-capable tool "signal_landscape"RecommendationScope tools to the minimum needed.
tool: farm_domain
The server advertises open-world / broadly-scoped capabilities.
no annotations on write-capable tool "farm_domain"RecommendationScope tools to the minimum needed.
tool: (server instructions)
The description pairs a fetch imperative with a hardcoded external URL, letting the server relocate its real instructions off-metadata and mutate them after review. Fires only when a fetch verb and a URL co-occur, so benign documentation links do not.
"Boolsai Scan" (https://boolsai.ai/mcp), "Boolsai Directory" (https://directory.boolsaRecommendationDo not direct the model to fetch and act on external URLs; treat linked content as untrusted.
Vantaj uptime monitoring via MCP — manage monitors, heartbeats, incidents, and status pages.
Unified gateway to Algeria's TKAWEN ecosystem: commerce, certification, and AI tools.
Provides access to the Cohereon Doctrine AI safety framework with governance components, tiered access, and agent onboarding.
Agentic rails for complex workflows with receipts, fees, and MCP tool access.
Structural TC39 spec lookup for ECMA-262 and ECMA-402 in AI agents, SHA-pinned and offline-first.
Structural TC39 spec lookup for ECMA-262 and ECMA-402 in AI agents, SHA-pinned and offline-first.
Live signal feed: events fired in the last N days (default 7). Returns each event with the predicted α range based on its event type's historical performance. Use this to surface 'what should I be loo
Deep dive on a single event: full diff (added/removed values), surrounding price action (-3D to +14D), predicted vs actual α, links to wayback comparison. Use this to investigate a specific event flag
Scan a URL as it appeared on a historical date via the Wayback Machine. Uses intel.boolsai.ai against the wayback-wrapped URL. Returns the same JSON shape as Boolsai Scan but for a historical snapshot
All events fired on a single ticker, plus price action timeline. Use this to investigate one company's pattern (e.g. 'show me everything we caught on NFLX').
Run an SPY-benchmarked backtest on the WAYBACK historical event dataset (2+ years, 13K events) instead of the recent live event dataset (2 months, 1.7K events). Much bigger samples for statistical con
Week-by-week wayback diff timeline for one domain. Returns every detected stack change (additions / removals) with week date. Use this to see when a vendor was added/removed historically, e.g. 'when d
ONE-SHOT cross-signal sweep. Computes α-vs-SPY stats simultaneously across event_type, detector, diff_field, severity, AND co_occurrence dimensions — returns the full landscape in a single response. U
Compare two signal patterns side-by-side. e.g. 'how does PRICING_TIERS_ADDED compare to VENDORS_DETECTED_CHANGED on the live dataset?' Returns α, %pos, sample size, worst/best trades for each, plus de
Bulk-farm a domain's historical wayback snapshots into our index. Use this when you need backtest history on a domain we haven't already farmed (i.e. wayback_backtest / domain_timeline return no data