The Research Report template makes your AI answer one research question with a sourced report: state the question precisely, scope the sub-questions, gather findings with sources and reliability, then synthesize an answer that separates fact from inference. You approve the scope before gathering and the report before it is shared.
6 steps · 2 approvals · 1 loop back
example · the steps are this template's real steps
For founders, analysts and marketers who research markets, competitors or options and need every claim tied to a source.
It is a stored workflow, not a prompt. Your AI client receives one step at a time from ConvOps over MCP, does the work, and advances. The engine holds the run at each of its two approvals until a person says yes.
6 steps, in this order. Below each one is the instruction your AI receives at that step, exactly as the template stores it.
State the research question precisely before anything else.
Capture:
A fuzzy question produces a fuzzy report. If the operator's ask is broad, narrow it with them before advancing.
Save the question and its context as a task note with task_notes_add. The next step (Scope) reads this note.
Read the task notes with task_notes_list: the note from the Question step contains the question and its context. Work from it.
Scope the research. What good looks like:
Save the scope as a task note with task_notes_add. Then ask the operator to approve the scope before gathering begins. This is a checkpoint: scope drift wastes the most time in research, so the AI cannot advance until you approve.
no rejection path · the run moves on only when approved
Read the task notes with task_notes_list: the note from the Scope step contains the sub-questions and boundaries. Work from it.
Gather evidence against each sub-question. For every finding capture:
Stay inside the approved scope. Note where evidence is thin or conflicting rather than papering over it.
Save the gathered findings and sources as a task note with task_notes_add. If it exceeds 10,000 characters, split across notes or store where the operator keeps documents and record the location. The Synthesize step reads this note.
Read the task notes with task_notes_list: the Gather note holds the findings and sources, and the Scope note holds the sub-questions to answer. Work from them.
Write the report:
Do not assert beyond the evidence. Distinguish fact from inference.
Save the full report as a task note with task_notes_add. If it exceeds 10,000 characters, split across notes or store where the operator keeps documents and record the location. The Final Review step reads this note.
Read the task notes with task_notes_list: the Synthesize note holds the report, the Gather note holds the sources, and the Question note holds what was asked. Work from them.
Review the report:
Ask the operator to approve the report. This is a checkpoint: once shared the report informs decisions, so the AI cannot advance until you approve. If you reject, the workflow returns to Synthesize.
rejected → back to Synthesize
The research report is approved and ready to share. Confirm the final report, with its sources, is saved as a task note, then close out the task. This note is the named input for a Decision Memo workflow if a decision follows.
Two steps wait for a person. You answer in the chat; the engine records the verdict and the audit row records who gave it.
Scope drift wastes the most time in research. You approve the sub-questions, the boundaries and the shape of the report before gathering begins.
what you review
no rejection path · approve to continue
Once shared, the report informs decisions. The AI checks it answers the question and every claim has a recorded source. Reject it and the run returns to Synthesize.
what you review
reject → back to Synthesize
gates the engine checks without you
Read from the template's own steps: the tools they call, the files they write, and whether they touch code.
Connect your AI client first, then install Research Report from the catalog in the ConvOps app. See how to connect a client
An installed template is a workflow in your workspace. Ask your AI to change it in plain words; the edit lands on the one step that needs it.
new chat
more example prompts
Gather records each finding with its source, reliability and caveat, and notes where evidence is thin or conflicting. Synthesize states uncertainty honestly and does not assert beyond the evidence.
ConvOps does not run models or browse. Your AI client does the research with whatever tools it has, and the workflow structures the question, scope, sources and review.
The Complete step saves the final report with sources as a task note, which is the named input for a Decision Memo workflow if a decision follows.
Any MCP client connected to ConvOps. Verified clients: Claude Code, Claude (desktop, web, Cowork), OpenAI Codex, ChatGPT, Cursor, OpenCode. It needs no repository, so a plain chat client works.