prepare_research
FBMCPF-263: deterministically assemble a research REQUEST packet for a ticket BEFORE implementation (no model calls). Returns the questions to answer — how to execute (approaches + tradeoffs), prior art IN THIS repo (files/tickets), comparables/competitors, risks/invariants — plus local sources t...
This record as markdown: /tools/ai-featureboard-featureboard/prepare-research.md
What prepare_research does on FeatureBoard
AI agents call prepare_research to retrieve information from FeatureBoard without modifying anything. It is typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
Why prepare_research is rated Low
The tool assembles and returns a research packet without executing code, modifying data, or making external calls. It is a preparatory/query operation that reads and collates existing information. The mention of 'saveInstruction' suggests it instructs an orchestrator to save via another tool (add_kb_doc), but this tool itself does not perform the write — it only returns instructions.
From the tool's definition deterministically assemble a research REQUEST packet... BEFORE implementation (no model calls). Returns the questions to answer... plus local sources to seed from
Attacks that exploit this kind of access
The rule that runs prepare_research safely
PolicyLayer is an MCP gateway: it sits between your AI agents and FeatureBoard, and checks every tool call against a rule you set before the call runs. Nothing changes on the server itself. For prepare_research, this is the rule to start with:
prepare_research is read-only, so it stays allowed. Everything else on the server is denied unless you say otherwise.
The button opens the PolicyLayer dashboard: create your workspace, connect FeatureBoard, apply this rule, and every prepare_research call is checked against it from then on.
Questions about prepare_research
FBMCPF-263: deterministically assemble a research REQUEST packet for a ticket BEFORE implementation (no model calls). Returns the questions to answer — how to execute (approaches + tradeoffs), prior art IN THIS repo (files/tickets), comparables/competitors, risks/invariants — plus local sources to seed from (matching KB docs, docs/ paths, code hints, and prior-art hits from the local lexical RAG, FBMCPF-264), a deliverable spec (a collated markdown brief ≤ ~150 lines), a saveInstruction (orchestrator saves the returned brief via add_kb_doc as research/<ticket> so getWorkPacket auto-attaches it as researchBrief), and a suggested cheap model (haiku for effort:low/medium, else sonnet). When the research phase resolves OFF (config researchOnIntake:false or a research:off label) returns { skip:true, reason }; a research:on label forces it on. It is categorised as a Read tool in the FeatureBoard MCP Server, which means it retrieves data without modifying state.
Register the FeatureBoard MCP server in PolicyLayer and add a rule for prepare_research: allow, deny, rate-limit, or require approval. Point your MCP client at the PolicyLayer proxy URL and the rule is enforced on every call, before it reaches FeatureBoard. Nothing to install.
prepare_research is a Read tool with low risk. Read-only tools are generally safe to allow by default.
Yes. Add a rate_limit block to the prepare_research rule in your PolicyLayer policy. For example, setting max: 10 and window: 60 limits the tool to 10 calls per minute. Rate limits are tracked per agent session and reset automatically.
Set action: deny in the PolicyLayer policy for prepare_research. The AI agent will receive a policy violation error and cannot call the tool. You can also include a reason field to explain why the tool is blocked.
prepare_research is provided by the FeatureBoard MCP server (https://github.com/valentil/featureboard-mcp/releases/download/v0.7/featureboard.plugin). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
More on FeatureBoard, and thousands of servers like it.
This server
Across the catalogue