How to use the Build Competitive Brief agent skill

What is an agent skill and what does this one do?

An agent skill is a folder containing a SKILL.md file and optional reference files. When it is installed, the agent can load it and follow its workflow when a request matches the skill's description. You do not need to run it yourself; you ask for the outcome in plain language.

This skill is for competitor or vendor comparisons, market-landscape analysis, battlecards, objection packages, positioning briefs and account-specific competitive views. It resolves the comparison anchor, gathers evidence in a set authority order, builds a dossier for each competitor and then writes the brief.

• Default output: a self-contained HTML document plus a concise chat summary.
• Chat or Markdown output only when you explicitly ask for it.
• Facts are kept separate from field signals and inference.
• Each competitor is kept as its own profile rather than merged into one generic story.
• Matrices compare at most five competitors unless you ask for more.

Install the build-competitive-brief skill

Download the skill ZIP from this page and unzip it. For a personal install, place the unzipped folder in ~/.claude/skills/. To make it available only inside one project, place it in that project's .claude/skills/ folder.

If you use the Claude apps instead, upload the ZIP under Settings → Capabilities → Skills. Use whichever route matches how you work; the skill's content is the same.

• Personal install: ~/.claude/skills/
• Project install: .claude/skills/ inside your project folder
• Claude apps: Settings → Capabilities → Skills, then upload the ZIP

What files come with the skill?

The package contains SKILL.md, which holds the workflow and guardrails, plus three reference files the agent loads only when they matter, and an agents/openai.yaml file.

• SKILL.md: the normal research flow, HTML-by-default behaviour and evidence rules.
• references/output-contract.md: exact section order, including compressed brief variants.
• references/request-schema.yaml: structured input, mode normalisation and matrix limits.
• references/source-priority.md: how to rank sources when they conflict.
• agents/openai.yaml: agent configuration file included in the package.

How to trigger the skill with natural requests

Describe the outcome you want. The skill's description covers competitor or vendor comparisons, market landscape analysis, battlecards, objection packages, positioning briefs and account-specific competitive views. Name the competitors, account or product category where you can, and use your own wording for objections.

If you give only a product, workflow, segment or market category with no competitor names, the skill will discover likely competitors and label its uncertainty. It asks you to confirm any required anchor it inferred from sources before doing broad research.

• "Build a battlecard for MyCompany against Competitor Alpha for a mid-market deal."
• "Give me objection handling for when a buyer says Competitor Alpha is cheaper."
• "Make a positioning brief comparing MyCompany with its main rivals in the workflow automation category."
• "I have an account review with ExampleCorp next week. Prepare defensive prep on competitors that might show up."

What to provide for better results

The skill works with material you supply alone. Your own notes, decks, call transcripts, prior briefs and linked documents are the strongest inputs because they rank first in its source order. Connected sources such as a CRM, knowledge base, meeting transcripts, internal messaging or sales intelligence are optional and are referenced through placeholders in the skill.

Public research needs internet access. The skill checks current competitor claims against official product, pricing, launch, customer-story, investor and regulatory sources, and uses secondary coverage only when primary sources are insufficient.

• Give the exact competitor names and the seller's product name.
• Paste or attach your own notes or battlecards if you have them.
• Say whether the output is for an account, a general market view or an objection.
• Connected CRM or transcript access is optional; the skill still works without it.

Safety notes: network access and connected data

According to the skill's package details, it runs no scripts but does use the network. Its public research means the agent will access the internet to read competitor sites and public material.

If a connected source is enabled, the agent may read it. Do not connect confidential CRM or deal data if you do not want an AI agent to reach it. The brief also relies on public sources and inference, so it is not a substitute for verified, audited due diligence for legal or investment decisions.

• Runs scripts: no.
• Uses network: yes, for public research.
• Connected CRM, transcripts and messaging may be accessed when enabled.
• Check the Sources section of every brief before sharing it.

Tips for a useful brief

Ask for the narrowest shape that fits your need. An objection request focuses on objection handling and proof points. An account overlay adds an Account Implications section and labels directional inference clearly. Account-scoped briefs that have no direct evidence of a competitor in the deal are labelled Defensive prep.

Read the What Not To Overclaim section before you use any line in front of a buyer. Field signals are treated as directional until corroborated, so treat them that way too.

• Use objection or account_overlay wording when that is what you need.
• Check that each competitor section has its own evidence.
• Verify any pricing or launch claim against the linked source before you repeat it.