bmad-deep-recon

Research a topic to support a decision, three ways: draft a research prompt for the user to run in their own tool (ChatGPT, Gemini, Grok, Perplexity, …), turn a finished research report into a short summary with cited sources that other skills can use directly, or run the research here with parallel web searches. Built-in research types: market, domain, technical, competitive, user-voice, academic-lit; also supports choosing between candidates, and custom types via overrides. Use when the user says "deep recon", "research this", "draft a research prompt", "process this research report", "market research", "domain research", "technical research", "competitor research", "literature review", or "help me choose between"

Quality

89%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Low

Low-risk findings worth noting

Quality

Content

78%Weight 40%Scale 1-5

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

The SKILL.md body functions as a well-structured overview that points to a real one-level reference layer, gives concrete commands and a headless JSON template, and explicitly assumes Claude's research competence. It is slightly prose-heavy in the Activation section and relies on external project scripts, which keeps actionability and conciseness just below top marks.

Suggestions

Tighten the On Activation prose (steps 1-5) into terser bullet form to cut tokens without losing the resolve_customization/resolve_config invocation sequence.

Inline a brief validate-then-proceed checkpoint for the Run plan gate (e.g. 'plan gate: halt until plan confirmed; on rejection revise and re-gate') so the feedback loop is visible in SKILL.md rather than only in verification.md.

Clarify which scripts ship with the skill bundle (scripts/recon_kit.py) versus which are project-level (_bmad/scripts/*) so the actionability of each command is unambiguous.

Dimension	Reasoning	Score
Conciseness	The body is efficient and assumes Claude's competence ("You already know how to research; the pack is where this harness is opinionated"), avoiding basic-concept padding, though the Activation prose is somewhat dense and could be trimmed further.	4 / 5
Actionability	Concrete executable commands are present (`uv run .../resolve_customization.py --skill ... --key workflow`, `uv run .../memlog.py --type <...>`, `uv run scripts/recon_kit.py tally ...`) plus a copy-ready headless JSON template, with only minor gaps around external project-level scripts.	4 / 5
Workflow Clarity	Sequencing is clear: a numbered 5-step Activation flow, an Intent routing table, and a Headless Mode path, with the plan gate flagged as "the one hard stop" and verification routed to a dedicated reference; validation feedback loops live in verification.md rather than inline.	4 / 5
Progressive Disclosure	The body is a concise overview that signals one-level-deep references via an Intent table mapping each intent to its reference file (run.md, draft.md, process.md, finalize.md, verification.md, synthesis.md, selection.md, lifecycle.md), all of which exist in references/; navigation is easy and nothing is deeply nested.	5 / 5
	Total	17 / 20 Passed

Description

100%Weight 40%Scale 1-5

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

The description is dense, specific, and self-contained: it names three concrete services, lists research types, and provides an explicit multi-phrase trigger clause answering both what and when. No fluff or over-claims detected, and the third-person voice is maintained throughout.

Dimension	Reasoning	Score
Specificity	Lists three concrete actions ("draft a research prompt", "turn a finished research report into a short summary with cited sources", "run the research here with parallel web searches") plus six named research types, giving comprehensive coverage rather than vague domain naming.	5 / 5
Completeness	It explicitly answers both what (three research services plus built-in types and custom overrides) and when (concrete "Use when the user says..." trigger phrases), matching the anchor for explicit what-and-when with concrete triggers.	5 / 5
Trigger Term Quality	An explicit "Use when the user says" clause enumerates ten natural phrases users would actually say ("deep recon", "research this", "market research", "literature review", "help me choose between") with good synonym coverage.	5 / 5
Distinctiveness Conflict Risk	It carves a clear niche (decision-grade research director) with distinct triggers like "deep recon" and "process this research report", minimizing overlap with generic search or analysis skills.	5 / 5
	Total	20 / 20 Passed

Validation

100%

Warnings & errors only

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository: bmad-code-org/BMAD-METHOD
Path: src/core-skills/bmad-deep-recon/SKILL.md
Commit: 05bfbd4

Reviewed: 15 days ago

Table of Contents

Discovery Implementation Validation

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.