advisor-profile
Cold-start interview that builds and maintains company-context.md — the persistent company/advisor context profile every C-suite advisor skill (ceo-advisor…
Skill by Anthropic (claude-for-legal, adapted)
Install
Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.
npx agentmarket install skill/bharatnain-advisor-profile
Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.
Get the weekly diff
The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.
About
Use cases: finance, legal
Reviews
Write a reviewNo reviews yet. Installed it? Reviews come from verified installers only. How reviews work
Permissions
What this skill can do and what it asks for, from its SKILL.md. Checked Oct 25, 2026.
- Tools
- No tools found
- Secrets
- None asked for
- Install scripts
- None found
Network hosts it talks to 1
codeload.github.com· found in install command
Files it touches
~/.claude/skills/advisor-profile
What's inside
- Skill name
advisor-profile- Description
- Cold-start interview that builds and maintains company-context.md — the persistent company/advisor context profile every C-suite advisor skill (ceo-advisor, cfo-advisor, cmo-advisor, cpo-advisor, cro-advisor, general-counsel-advisor) reads before advising. Captures business model, stage, org, financial posture, GTM motion, and legal/regulatory surface once so advisors stop re-asking. Use on first use of any advisor skill, when company-context.md is missing or full of placeholders, or when the user says set up my company profile, onboard me as an advisor, update my company context, or the advisors keep asking the same questions. Supports quick (2-min) or full (15-min) depth, pause/resume, --redo with diff, and --section partial re-interview.
- License
- Apache-2.0
- Metadata
- author: Anthropic (claude-for-legal, adapted) · domain: advisor-context · version: 1.0.0 · category: c-level · frameworks: cold-start-interview, company-context-profile
- Instructions
- 225 lines in SKILL.md · view on GitHub
Agent Skills spec issues
- files: referenced but missing: references/company-profile-template.md
Files (3)
- LICENSE.txt
- SKILL.md
- company-context-template.md · references/
SKILL.md (first lines)
# Advisor Profile (Cold-Start Interview) > Adapted from [anthropics/claude-for-legal](https://github.com/anthropics/claude-for-legal) > `commercial-legal/skills/cold-start-interview` + `references/company-profile-template.md` > (Apache-2.0 — see `LICENSE.txt` in this directory). Genericized from a legal practice > profile to a company/advisor context profile for this toolkit's C-suite advisor skills. Runs the cold-start interview and writes the **company context profile** — the file the C-suite advisor skills (`ceo-advisor`, `cfo-advisor`, `cmo-advisor`, `cpo-advisor`, `cro-advisor`, `general-counsel-advisor`, `executive-mentor`) read before doing anything. First run writes the profile; subsequent runs with `--redo` re-interview and show a diff before overwriting. ## Where the profile lives Two locations, checked in this order by every advisor skill: 1. **Project-local:** `company-context.md` in the project root. The advisor skills already read this file if it exists. Use it when the profile belongs to one repo/venture. 2. **Global (version-independent):** `~/.claude/plugins/config/bharats-claude-toolkit/company-context.md`. This path is deliberately **outside any plugin cache or versioned install directory**, so it survives plugin updates and reinstalls. Use it when one company context should follow the user across projects. Default behavior: write the global profile, then offer to copy it into the current project as `company-context.md` (project-local wins when both exist). Create parent directories as needed. The section scaffold lives at [references/company-context-template.md](references/company-context-template.md). ## Instructions 1. **Check current state:** Read the project-local `company-context.md`, then the global path. Classify: - **Neither exists** → start a fresh interview. - **Contains `<!-- SETUP PAUSED AT: -->`** → greet the user and offer to resume from that section. - **Contains `[PLACEHOLDER]` or `[Your Company Name]` markers but no pause comment** → the template was never completed; offer to start fresh or resume from wherever the placeholders begin. - **Populated (no placeholders, no pause comment)** → already configured; skip unless `--redo` or `--section <name>` was asked for. If populated and the user still asked to run setup, ask: "Looks like you're already set up. Want to re-run the interview? This will overwrite the profile (I'll show you a diff first)." 2. **Show the depth fork** (quick vs. full — see below) and wait for the pick. 3. **Follow the interview script below** at the chosen depth. 4. **Ask for seed docs:** pitch deck, latest board deck, financial model, website/about page, org chart — whatever exists. Accept file paths, links, or pastes. Seed docs beat memory: extract facts from them instead of making the user re-type. 5. **Write the profile** per the template structure. Use the user's own words where possible. 6. **Show summary + propose next steps:** "Here's what I heard — the profile is written. What did I get wrong?" Then offer a test run: "Want to ask one of the advisors something and see how the context lands?" ## Flags ### `--redo` Full re-interview of an already-configured profile. Before overwriting, show a diff between the current profile and the newly drafted one, section by section, and get an explicit yes. Never silently overwrite a populated profile. ### `--full` Upgrade a quick-start profile to the full interview. Re-ask only the sections still marked `[DEFAULT]` or `[PLACEHOLDER]`; keep everything the user already answered. ### `--section <name>` Partial re-interview of exactly one section, leaving the rest untouched (adapted from the upstream `--side` mechanic). Valid names map to the profile's sections: | `--section` | Re-interviews | |---|---| | `business-model` | What you sell, to whom, pricing/revenue model, stage and funding | | `org` | Team size, leadership, who owns what, hiring posture | | `financials` | Revenue band, growth, burn/runway posture, margin structure, fundraising plans | | `gtm` | Sales motion, ACV band, sales cycle, marketing engine, ICP | | `legal` | Jurisdictions, regulated data, contract posture, counsel access, contract playbook | | `risk` | Risk appetite, escalation/decision rights, the one thing that keeps leadership up at night | Do NOT re-ask the other sections — they are already populated. Use this when (a) something changed (a fundraise closed, a pivot, a new GTM motion), or (b) a section was skipped at setup and an advisor skill just hit the gap. ## Purpose You are meeting this company for the first time. Your job is to learn how *this* company actually works — not how startups work in the abstract — and write what you learn into a living context profile that every advisor skill reads before it advises. The user should leave this conversation feeling like they just onboarded a sharp new chief of staff who asked exactly the right questions. They should never see a YAML config file. They should see a document about their company that they can edit in plain English. **Why this matters.** Every advisor skill reads this profile. A generic profile gives you generic advice — default benchmarks, a default stage assumption, and recommendations that feel like they were written for someone else's company. The more specific the answers — your real runway, your real ACV, your real deal-breaker — the more the advisors' output will feel like it was written by someone who works there. **Fresh profile from explicit inputs only.** The interview's inputs are the user's typed answers and documents they explicitly share. Do not pull from ambient context, prior sessions, or user memory to fill in gaps. If something relevant surfaced earlier in the current conversation, ask before using it. ## Before the interview starts Show the fork-first preamble — 3-4 short lines, no longer: > **This interview writes the company context profile that all the C-suite advisor skills read** (CEO, CFO, CMO, CPO, CRO, GC). Answer once; stop re-explaining your company to every advisor. > > **2 minutes** gets you stage, business model one-liner, GTM motion, and role — plus working defaults for everything else, each marked `[DEFAULT]`. **15 minutes** adds your real org shape, financial posture, ICP and sales cycle, legal/regulatory surface, escalation rules, and the facts extracted from your pitch deck or board materials. > > Quick or full? (Upgrade any time with `--full`.) Wait for the user's pick before showing anything else. **Quick start path:** ask only Part 0 and the business-model basics (stage, one-liner, GTM motion). Write the profile with `[DEFAULT]` markers on everything else. Close with: "Done. The advisors work now. When an advisor's output feels off, that's usually a default you should tune — it'll tell you which. Run the interview with `--full` anytime, or `--section <name>` to redo one part." **Full setup path:** the whole interview below. ## Interview pacing - **Assume the answer exists somewhere.** When a question asks for information that's probably written down — company description, org chart, financial model, ICP doc — prompt for a link or a paste before asking the user to type it from memory. "Paste a link or a doc, or give me the short version" is the default ask for anything longer than a sentence. - **Batch size — count subparts.** Never more than 2-3 *answerable prompts* per turn, counting subparts. One question with 5 subparts is 5 questions. The test: can the user answer without scrolling? - **Ask and wait.** When a question needs a typed answer, say so and wait. Do not move on until the user responds. - **For uploads and seed docs:** "Paste the contents, share a file path, or say 'skip for now.' If you skip, I'll flag the gap in the profile so you can fill it later." Then actually wait. - **Before writing the profile:** list any questions that were skipped or answered with placeholders. Say: "Before I write the profile, here's what's still open: [list]. Want to fill any of these now, or leave them as placeholders?" Then wait. **Never** write a profile with silent gaps — every placeholder should be a deliberate skip, not a question that scrolled past. - **Pause and resume.** Tell the user up front: "If you need to stop, say 'pause' and I'll save your progress." On pause, write a partial profile with a `<!-- SETUP PAUSED AT: [section name] — run the advisor-profile interview to resume -->` comment at the top and `[PENDING]` markers (distinct from `[PLACEHOLDER]`) on unanswered fields. When setup re-runs and finds a paused profile, greet the user: "Welcome back. You paused at [section]. Your earlier answers are saved. Pick up where we left off, or start over?" Do not re-ask questions already answered. - **Sanity-check specific claims as they come up.** When the user states a specific number, metric definition, or regulatory claim you can sanity-check (e.g., "our net revenue retention is 200%", "GDPR doesn't apply to us"), and it conflicts with what they've pasted or with your understanding, surface it before writing it in: "You said X; the deck says Y — which goes in the profile?" A wrong fact written into the profile propagates into every future advisor output. ## The interview ### Opening > I'm going to be the shared memory behind your advisor team. Before any of them advise you on anything, I want to learn how your company actually works — not generic startup patterns, but *your* model, *your* numbers posture, *your* deal-breakers. > > This takes about fifteen minutes. I'll ask a few questions, then I'll ask you to point me at a pitch deck or board deck so I can pull facts from documents instead of making you type them. ### Part 0: Who's using this > Who'll be using the advisor skills day to day? > > 1. **Founder / CEO** — the buck stops with you. > 2. **Exec team member** — you own a function (finance, product, marketing, revenue, legal) and report to the CEO. > 3. **Operator / chief of staff** — you prepare analysis for decision-makers. > 4. **Advisor / consultant** — you advise this company but don't work there. Record the role — it shapes voice (decisions vs. briefing material) and how strongly advisors should push "your decision" framing. If the answer is 4, note that outputs should be framed as material for the operating team, not decisions. ### Part 1: Business model and stage (2-3 minutes) **What does [the company] do?** This is the single most important context — a PLG SaaS playbook, a marketplace playbook, and a services-firm playbook are completely different. You don't have to type it out: paste a link to the website, the about page, or the latest pitch deck, and I'll extract what I need. Or give me the one-sentence version: what you sell, to whom, and how it's priced. **Who are you?** - Company name and entity type (Delaware C-corp? LLC? Something else?)
From bharatnain/bharats-claude-toolkit, Apache-2.0.
More from bharatnain/bharats-claude-toolkit
- 0.587score
agent-wrap-up
bharatnain
Close out (or revisit) a Claude Managed Agent (CMA) build from /launch-your-agent — refresh the overview page, recap every CMA primitive the user now owns…
SkillAFreshness: A★ 1Claude Code +9 - 0.587score
3-statement-model
bharatnain
Complete, populate and fill out 3-statement financial model templates (Income Statement, Balance Sheet, Cash Flow Statement) .
SkillAFreshness: A★ 1Claude Code +9 - 0.582score
audit-xls
bharatnain
Audit a spreadsheet for formula accuracy, errors, and common mistakes.
SkillAFreshness: A★ 1Claude Code +9 - 0.582score
api-idempotency-keys
bharatnain
Idempotency key design -- UUID v4 generation, key storage TTL, 24h window convention, at-least-once vs exactly-once semantics, safe retry scope.
SkillAFreshness: A★ 1Claude Code +9 - 0.582score
api-rate-limiting
bharatnain
Rate limit design as consumer contract -- quota tiers, burst vs sustained limits, per-user vs per-app limits, fair-use policy, quota negotiation.
SkillAFreshness: A★ 1Claude Code +9 - 0.582score
api-webhook-design
bharatnain
Webhook registration, payload design, delivery guarantees (at-least-once), retry policy, ordering guarantees, fan-out patterns.
SkillAFreshness: A★ 1Claude Code +9 - 0.570score
board
bharatnain
Build and open the orchestrator board — what is waiting on you, what is happening now, in-flight beads work, sessions and subagents with model and tokens, the…
SkillAFreshness: A★ 1Claude Code +9 - 0.562score
ad-creative
bharatnain
When the user wants to generate, iterate, or scale ad creative — headlines, descriptions, primary text, or full ad variations — for any paid advertising…
SkillAFreshness: A★ 1Claude Code +9 - 0.562score
ai-seo
bharatnain
When the user wants to optimize content for AI search engines, get cited by LLMs, or appear in AI-generated answers.
SkillAFreshness: A★ 1Claude Code +9 - 0.557score
cfo-advisor
Alireza Rezvani
Financial leadership for startups and scaling companies.
SkillAFreshness: A★ 1Claude Code +9 - 0.557score
board-deck-builder
Alireza Rezvani
Assembles comprehensive board and investor update decks by pulling perspectives from all C-suite roles.
SkillAFreshness: A★ 1Claude Code +9 - 0.557score
ceo-advisor
Alireza Rezvani
Executive leadership guidance for strategic decision-making, organizational development, and stakeholder management.
SkillAFreshness: A★ 1Claude Code +9
Signals
- Stars
- 1
- Forks
- 0
- Open issues
- 1
- Last commit (this folder)
- Sep 24, 2026 (1 month ago)
- Last release
- Oct 22, 2026 (6 days ago)
From GitHub (fetched Oct 24, 2026).
Details
- Version
- 1.0.0
- License
- Apache-2.0
- First seen
- Oct 24, 2026
- Last checked
- Oct 24, 2026
- Source
- GitHubgithub-graphql
Cite this page
advisor-profile skill. Agent Mart. https://agentmart.dev/skills/bharatnain-advisor-profile (last updated 2026-10-24)