What does this skill do, and when should you use it?
This skill turns Claude into an Indian equity research analyst producing 1,500-3,000 word institutional-grade reports for companies listed on the NSE or BSE. It enforces strict anti-hallucination rules: every financial figure must come from a live fetched page (Screener.in, BSE filings, Trendlyne, Tickertape) and carry an inline [Source: URL, date] tag, or the report must state 'data unavailable'. Analysis covers fundamentals, forensic accounting red flags, technicals, peer comparison, multi-model valuation (DCF, EV/EBITDA, P/E re-rating) and TAM sizing. Output is a DOCX report with a SEBI disclaimer plus a markdown summary in chat. A mandatory self-audit checklist (price snapshot, Bull/Base/Bear table, split verdict) gates delivery.
Resolves a company name to an NSE ticker, fetches the current market price via web search (Google Finance card, Tickertape, 5Paisa — explicitly rejecting Screener.in and Yahoo Finance as price sources), then fetches historical financials, shareholding, concall transcripts, news, technicals, peer valuations and analyst consensus. It runs analysis frameworks (DuPont ROE decomposition, ROIC-WACC spread, forensic accounting checks), derives a target price with sensitivity tables, self-audits for deliverability, and saves a .docx to /mnt/user-data/outputs/ alongside a pasted markdown summary.
- An individual investor wanting a Buy/Sell/Hold call on a specific NSE/BSE stock, e.g. RELIANCE or HDFCBANK.
- A portfolio holder auditing multiple positions at once ('Analyze my holdings: HDFCBANK, INFY, TCS, RELIANCE').
- An investor comparing competitors (Tata Motors vs Maruti Suzuki) with a valuation differential table.
- An event-driven analyst researching corporate-action impact, e.g. HDFC Bank post-merger.
- A small-cap/SME investor researching thinly covered BSE names like Macfos Ltd (BSE: 543787).
How do you install this skill?
- Core price fetching relies on the Google search result card — unusable from mainland China, and snippet-card prices can themselves be delayed or inaccurate.
- The entire pipeline scrapes third-party pages (Screener/Trendlyne/Tickertape); layout changes or anti-bot measures will break the skill with no documented degraded mode.
- README performance claims (98% accuracy, 50+ stocks tested) have no committed supporting evidence — treat as marketing.
- Reports produce buy/sell levels, stop-losses and targets — borderline investment advice; users must independently verify data and consult licensed advisors.
- No tests, no CI, empty examples link, unverified publisher — maintenance sustainability is uncertain.
- Network access
- Local filesystem
web_search / web_fetch toolsdocx output capability
Clone the repo (git clone https://github.com/vishalmdi/india-equity-report-skill.git), copy the india-equity-report folder into your Claude skills directory (cp -r india-equity-report /path/to/your/claude/skills/user/), then restart your Claude Code session. The repo documents this for Claude Code; other platforms are undocumented.
tmp="$(mktemp -d)"
git clone --depth 1 https://github.com/vishalmdi/india-equity-report-skill.git "$tmp"
mkdir -p ~/.claude/skills
cp -R "$tmp/india-equity-report" ~/.claude/skills/
rm -rf "$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:
- Analyze Reliance Industries (RELIANCE)
- Should I buy ITC at current levels?
In Claude Code, simply prompt with a stock name or ticker, e.g. 'Analyze Reliance Industries (RELIANCE)' or 'Should I buy ITC at current levels?'. The skill triggers on stock-analysis phrases, fetches live prices and financials automatically, runs its quality checklist, and presents the DOCX report via present_files. For batch analysis, list multiple tickers in one prompt.
What are this skill's strengths and limitations?
- Seven-layer anti-hallucination guardrails with mandatory source citations on every figure.
- Multi-dimensional analysis: fundamentals, forensic accounting, technicals, and multi-model valuation.
- Detailed source hierarchy with explicit warnings against unreliable price sources (Screener.in stale cache, Yahoo Finance delayed quotes).
- Mandatory pre-publish self-audit checklist (price, data integrity, formatting, docx).
- Works on large-cap, mid-cap and small-cap/SME stocks per the README's tested examples.
- Output quality depends on web search/fetch tooling available in Claude Code — no offline mode.
- Relies on scraping third-party sites (Screener.in, Trendlyne, Tickertape) that may change layouts or block access.
- No independent test suite; the README's 98% accuracy and satisfaction stats are self-reported.
- License field on GitHub is NOASSERTION despite README stating MIT — verify before commercial use.
- Not registered investment advice — outputs are informational only and subject to a SEBI disclaimer.
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 |
|---|---|---|---|---|
| India Equity Report Skill this page | 45 · Not recommended | ★ 26 | 7mo ago | NOASSERTION |
| Tech Earnings Deep Dive Analyzer | 47 · Use with care | ★ 1.1k | 6mo ago | MIT |
| Taiwan Stock One-Click Financial Model & Investment Deck Generator | 45 · Not recommended | ★ 40 | 3mo ago | — |
| Nine-Chapter Equity Research Skill | 63 · Recommended | ★ 462 | 2mo ago | MIT |
| Buffett Investment Thinking System | 53 · Use with care | ★ 823 | 6mo ago | — |
How did FollowSkills review this skill?
Least privilege, no sensitive-data operations, no external side effects; strong mandatory source-attribution and SEBI disclaimers; data flow is transparent. Deductions: no user-confirmation mechanism, no rollback design, and the core CMP fetch relies on search-engine snippet cards and scraping third-party pages, an uncontrollable MITM-prone path; attribution execution unverifiable statically.
Workflow is structurally complete (Steps 0-5, QC checklist, anti-hallucination rules, explicit 'data unavailable' fallback). Deductions: static ceiling with no tests, CI, or committed sample reports; key paths depend entirely on third-party page structures and Google search cards — any layout change or anti-bot measure breaks the skill, with no documented degraded-mode guidance.
Trigger phrases and audience are clearly declared. Deductions: capability boundaries (behavior with no web tools or blocked sites) undeclared; core function depends on the Google search card, unreachable from mainland China; reachability of Tickertape/5Paisa/Trendlyne from China is unassessed and no alternatives are offered.
Good layered docs (SKILL.md + three references), MIT license with financial-use disclaimers, troubleshooting section. Deductions: placeholder links in README ([your-username], [email]) indicate rough maintenance; referenced examples/ directory not evidenced; no changelog; version inconsistency (1.0.0 badge vs v1.1 in SKILL.md); publisher unverified.
Instructions should steer a complete, source-cited report with marginal value over manual research. Deductions: 'battle-tested' cases and 98% accuracy claim have no committed evidence; the examples download link points to nothing; static review cannot confirm outputs are directly usable.
Mandatory per-figure [Source: URL, date] tags are good design. Deductions: repo-level evidence is thin — no tests, no CI workflows, no sample report files; performance stats (98% accuracy, 50+ stocks, five-star satisfaction) are unverifiable marketing claims.
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 →