Description
BMad Deep Recon is a research director for decisions. Its job is to frame research worth running, then turn the results into an artifact a project can use. It works three ways that can be combined. It can draft a deep-research prompt for you to run in your own tool, such as ChatGPT, Gemini, Grok or Perplexity. It can process a finished report into a short, cited summary that later skills read. Or it can run the research itself through parallel web searches. The skill follows two standing rules. It does not draw conclusions from training data alone, so every claim must be backed by evidence gathered or imported during the run. And project context, such as briefs or PRDs, shapes which questions get asked but is never treated as evidence. Use it when a choice needs research behind it, such as entering a market, picking a stack or committing to a domain. Like other BMad skills, it reads its settings from the project's _bmad/ folder. It is part of the BMad Method project on GitHub, which has about 54,000 stars. The listing shows it as free. Its security audits show passes from Gen Agent Trust Hub and Socket, and a warning from Snyk.…
Highlights
- Drafts a research prompt to run in your own tool
- Turns a finished report into a cited summary
- Runs research through parallel web searches
- Every claim must trace to gathered evidence
- Built for market, stack and product decisions
Ratings & Reviews
No ratings yet
Be the first to rate bmad-deep-recon.
People also visit
bmad-deep-recon FAQ
What does Deep Recon do?
It frames a research question around a decision, then either drafts a research prompt, processes a finished report into a cited summary, or runs the research through parallel web searches.
Which decisions is it for?
Choices such as entering a market, picking a stack or scoping a product. The skill shapes the research around the decision.
Does it treat my project files as facts?
No. Briefs, PRDs and code shape what is asked, but they are not evidence. Every claim in the output has to trace to a source the run gathered or imported.
Can I use it with my own AI tool?
Yes. It can draft a deep-research prompt for you to run in ChatGPT, Gemini, Grok, Perplexity or another tool, then process the report you bring back.
How do I install it?
Run npx skills add https://github.com/bmad-code-org/bmad-method --skill bmad-deep-recon.
Is it free and checked for security?
The listing shows it as free. Its security audits show passes from Gen Agent Trust Hub and Socket, and a warning from Snyk.
What are the best bmad-deep-recon alternatives?
Popular alternatives to bmad-deep-recon include bmad-build, bmad-code-review and bmad-correct-course. See all bmad-deep-recon alternatives