Slatemark agent plugin
Check your risk before you take the trade.
This package connects a supported AI client to Slatemark: the remote MCP server for delayed market data, technical calculations, SEC filings, and macroeconomic research, plus bundled analyst and workflow skills. Your private journal and framework rules are optional context. Slatemark supplies factual tool results; your AI client creates the response.
The package follows the Agent Plugins 1.0 format
(https://agent-plugins.org/specification): plugin.json is the manifest,
mcp.json declares the Slatemark server, and skills/ holds one directory per
skill. mcp_config.json declares the same server for AI clients that read
that file instead. The package has no commands, hooks, agents, or executable
code. Install it with
your AI client's plugin install flow, pointed at this slatemark directory.
Slatemark generates the skills/ files when it publishes the package, so a
copy taken from Slatemark's source repository needs
python scripts/build_plugin_skill.py agent before it is installed.
Start here
Use the first Free research workflow at https://slatemark.ai/first-workflow:
Show AAPL's latest available quote and latest non-null daily RSI(14) and ATR(14). Keep quote and indicator timestamps separate, and include each tool's source plus available
as_of,fetched_at, anddata_qualityfields.
It needs no journal entry, brokerage link, paid plan, or provider key. It does require an authenticated, available Free connection, sufficient daily history, and an available market-data source. Missing values or source failures remain missing. For other research or recorded-trade tasks, ask for a focused skill by name:
catalyst-mapfor dated events around a name.regime-checkfor the cross-asset backdrop. Plus.pre-trade-brieffor a ticker and horizon.position-reviewfor a position you already hold. Plus.earnings-setupfor the next earnings event. Plus.post-mortemfor a trade you closed.
Each workflow uses only the data available on the user's plan, names any
unavailable section, and continues with the evidence it can retrieve. The
broader analytical-framework skill provides optional methodology alongside
these focused workflows. Its generated frontmatter carries the published skill
version and content hash.
Connection and plans
The package connects to https://slatemark.ai/mcp over Streamable HTTP.
Sign-in uses OAuth, which the Agent Plugins format leaves to your AI client.
There are no static credentials in the package and no token to paste. A
Slatemark account is required. Free includes one constrained MCP connection.
Plus adds current brokerage Account Data, available past booked activity
imported into the trade journal, newly available booked activity imported
periodically, short interest history from FINRA, additional active AI-client
connections, and a higher fair-use limit.
Boundaries
Brokerage, order, trading, and funds access is read-only. Slatemark never places, modifies, or cancels an order; every trading decision is yours. User-directed journal writes change only your Slatemark records; the Plus booked-activity imports described above add brokerage activity to your trade journal. Nothing here is personalized investment advice, and Slatemark is not a registered investment adviser, broker-dealer, or fiduciary.
Market data is approximately 15 minutes delayed and labeled with its freshness. Brokerage connections provide Account Data, not market data. Option-chain snapshots are delayed, available only where listed, contain no Greeks, and use indicative, non-executable marks.
Distribution status
Slatemark distributes this package only through the public
pdassoc-io/slatemark-plugin Git repository. Git availability is not a
listing in any AI client's plugin directory or marketplace. The #202
combined-service gate is complete; the package must continue to preserve the
reviewed read-only, no-embedded-inference product shape.
Privacy and support
Slatemark's Privacy Policy and Terms of Service cover the hosted service this package connects to. For support, email support@slatemark.ai.
License / use
© 2026 PD&A LLC. This package is published solely for use with the hosted Slatemark service. It is not open-source software. You may download, install, and privately modify this package to connect to your Slatemark account, but you may not redistribute it. See the LICENSE file for full usage rights and disclaimers.