What does this skill do, and when should you use it?
This skill supports quarterly portfolio reviews, annual planning, and decisions about where to deploy AI investment first. It reads quarterly updates, board materials, and financials, then evaluates each company’s data readiness, internal ownership, and ability to run a 30-day pilot. It identifies opportunities across back-office, revenue, and sector-dependent operations and ranks them by EBITDA impact, speed to value, and execution probability. The output is a one-page operating-partner brief with company decisions and reusable portfolio playbooks.
It first asks where portfolio materials are stored: connected MCP servers, a local folder, or uploaded PDFs, PowerPoint files, and Excel workbooks. For each company, it extracts sector, revenue, functional headcount, mentioned technology stack, and existing AI or automation initiatives, while checking remaining hold period and prior successful deployments. It applies three gates—data availability, management ownership, and a 30-day pilot—and identifies two or three leverage points, including estimated weekly FTE-hours saved and whether each needs off-the-shelf software or a light build. It then ranks all opportunities across the portfolio and produces top opportunities, replays, Go/Wait decisions, rejected ideas, and aggregate Year 1 versus Years 2–3 EBITDA contribution.
- An operating partner is preparing a quarterly portfolio review and must decide which companies receive AI attention first.
- A private-equity team is building its annual value-creation plan and needs to compare quick wins across portfolio companies.
- Management proposes an automation pilot, and the team needs to test whether clean data, an accountable owner, and a 30-day pilot are realistic.
- One portfolio company has a working invoice-processing setup, and the team wants to identify suitable follower companies.
How do you install this skill?
- Do not treat estimated EBITDA contribution, probabilities, or AI-investment priority as validated conclusions; qualified professionals should verify the data, assumptions, and calculations.
- Before using portfolio-company financial or board materials, add access-control, confidentiality, retention, and third-party MCP data-flow requirements.
- Reachability of overseas MCP services from mainland China is not established; confirm that local-file or upload workflows are available.
- Network access
- Local filesystem
- MCP Server
In Cowork, open Settings → Plugins → Add plugin, paste https://github.com/anthropics/financial-services, and select the private-equity vertical from the marketplace list. Alternatively, upload a zip containing plugins/vertical-plugins/private-equity/. The supplied source does not document a dedicated Claude Code install command for this individual skill.
tmp="$(mktemp -d)"
git clone --depth 1 https://github.com/anthropics/financial-services.git "$tmp"
mkdir -p ~/.claude/skills
cp -R "$tmp/plugins/vertical-plugins/private-equity/skills/ai-readiness" ~/.claude/skills/
rm -rf "$tmp"Generated from the source repository and skill path; it copies only this skill's folder. If the author's install steps above differ, follow those first. To scope it to one project, replace ~/.claude/skills with that project's .claude/skills.
How do you use this skill?
Once installed, send your agent any of these to trigger it:
- AI opportunity scan
- where should we deploy AI
- AI across the portfolio
- AI quick wins
Use a trigger such as "AI readiness", "AI opportunity scan", "where should we deploy AI", "AI across the portfolio", "AI quick wins", or "which portcos are ready for AI". Example: "Run an AI readiness scan across the portfolio. The quarterly decks are in /path/to/files; rank the top five opportunities and identify replays." Provide the quarterly updates, board materials, and financials, plus each company’s remaining hold period and any known successful AI deployment.
What are this skill's strengths and limitations?
- Uses explicit Go/Wait gates for data, ownership, and pilot speed.
- Ranks opportunities by EBITDA impact, speed to value, and probability rather than novelty.
- Looks for repeatable plays, fast followers, and shared-vendor leverage across companies.
- Prioritizes off-the-shelf tools, 30-day pilots, and hold-period urgency.
- Results depend on complete portfolio materials and accurate identification of a management owner.
- The skill does not include data connectors, valuation models, or automation scripts; users must provide an MCP connection, files, or uploads.
- The supplied source includes no test suite, platform coverage evidence, or reported deployment results.
- It provides a ranking framework but no standardized EBITDA estimation formula or tool-cost dataset.
How does this skill compare with similar options?
Side by side with related skills; every score comes from the same FSRS standard.
| Skill | FS score | Stars | Last updated | License |
|---|---|---|---|---|
| Portfolio AI Readiness Scan this page ✓ Anthropic · Official | 46 · Use with care | ★ 40k | 1mo ago | Apache-2.0 |
| PE Returns Sensitivity ✓ Anthropic · Official | 46 · Use with care | ★ 40k | 1mo ago | Apache-2.0 |
| Tax-Loss Harvesting Advisor ✓ Anthropic · Official | 49 · Use with care | ★ 40k | 1mo ago | Apache-2.0 |
| cuOpt Multi-Objective Explorer ✓ NVIDIA · Official | 56 · Use with care | ★ 3.6k | 11d ago | Apache-2.0 |
| Azure AI Foundry Fine-Tuning ✓ Microsoft · Official | 48 · Use with care | ★ 2.7k | 3mo ago | MIT |
How did FollowSkills review this skill?
The skill asks the user to identify the data location and offers local files or uploads, with no stated credential collection or automatic external action. However, it handles financial, board, and portfolio-company data without explicit least-privilege controls, data-flow disclosure, sensitive-data handling, result confirmation, isolation, or rollback guidance, so points are deducted.
The sequence, Go/Wait gates, single-company exception, and output structure are broadly consistent. It lacks scripts, tests, input validation, missing-data handling, reproducible calculation rules, and diagnosable abnormal-input feedback; under static calibration it cannot exceed 10, so points are deducted.
The audience, quarterly-review and planning scenarios, trigger phrases, and single-company versus portfolio modes are reasonably clear. Non-fit boundaries, precise input formats, sector variation, and Chinese-language behavior are unspecified; some MCP options may depend on overseas connectivity, although local-file and upload alternatives exist, so points are deducted.
The document uses progressive, stepwise organization and includes an output table, operating principles, limitations, and failed-pilot guidance. The skill itself lacks versioning, changelog, named maintenance responsibility, update path, installation/dependency notes, and FAQs; governance evidence is mainly repository-level README material, so points are deducted.
The workflow covers scanning, gating, portfolio ranking, replay identification, and a one-page deliverable that could be directly reviewed. However, EBITDA, probability, and time-saved estimates lack a standardized calculation method, outputs require substantial professional review, and static reading cannot verify output quality, so points are deducted.
The gates and ranking criteria are manually auditable, and repository-level CI includes plugin validation and secret scanning. There are no skill-specific key-path tests, representative outputs, third-party corroboration, or independently reproducible execution results, so only limited credit is justified.
Open a dimension to read why it scored that way
Evidence confidence：Low — Mostly static review, author material or a limited demo; useful for discovery, not high-risk decisions.
See the full review method →