# Sip Review

> Health check across all the user's SIPs and funds as a set — expense ratio earned, category-rank drift, manager changes, overlap creep, continue/pause/redirect per SIP. Use when the user asks to review their SIPs, whether to continue one, or for a periodic MF checkup. Read-only — never changes a SIP.

- Skill: `getlost01/sip-review` (Agent Skill)
- Install (CLI): `npx skillmds@latest add getlost01/sip-review`
- Raw SKILL.md: https://api.skillmd.com/api/skills/getlost01/sip-review/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Research & Search
- Author: getlost01 (https://skillmd.com/u/getlost01)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/getlost01/sip-review

---


# SIP Review

Read-only — never start, stop, pause, or modify a SIP.
`reference/READ-ONLY-POLICY.md` (hard rule) and
`reference/RESEARCH-STANDARDS.md` (MF peer framework, tool
availability, data efficiency, disclosure) apply. Where
`mutual-fund-analysis` deep-dives one fund on demand, this sweeps the
whole set on a cadence. Works over whichever broker(s) `BROKERS.md`
names active — resolve capabilities against
`reference/BROKER-CAPABILITIES.md`.

Groww's MCP returns no mutual fund data at all. If INDmoney is active,
its SIP-status and holdings capability (`mf_sips`, `indian_stocks_sips`,
`networth_holdings` — **unverified**, check the capability map's status
table) covers this skill directly; otherwise it runs entirely on
`PORTFOLIO-PLAN.md` plus the web. With no SIP register, no fund list, and
no active INDmoney, there is nothing to review — say so and offer
`portfolio-plan-builder` rather than producing a hollow sweep.

## Steps

1. **Assemble the set.** Funds held, units and cost basis, plus SIP
   amounts and dates: from INDmoney's holdings/SIP capability if active,
   otherwise from `PORTFOLIO-PLAN.md`'s holdings and **SIP register** —
   no other broker here exposes either. Empty or stale → offer
   `portfolio-plan-builder` rather than guessing. Flag any fund listed
   with no SIP row, and any SIP row with no holding. Note the as-of for
   every ₹ figure that follows — the register's `_Last reviewed:_` date
   when sourced from the plan, live if from INDmoney.

2. **Per-fund scorecard**, peer framework applied lightly across the set
   — full depth only for funds that flag:
   - **Rolling** returns vs. benchmark and category median across
     1Y/3Y/5Y, not one point-to-point window — is the expense ratio
     earned with actual alpha, net of that expense ratio? Tracking error
     instead, for index funds. Prefer INDmoney where active; otherwise
     web-sourced (AMFI / Value Research / factsheet), one date-anchored
     search per fund at most, cited.
   - Category-**rank percentile** drift across 1Y/3Y/5Y, not a single
     period: a former top-quartile fund sliding to middling for 2+ years
     running is a flag; one bad year isn't.
   - One date-anchored `WebSearch` for flagged funds or on a full annual
     review — manager change, mandate change, AMC issue.
     `fund-house-watch` covers AMC-level flags.

3. **Set-level checks:** overlap creep (funds converging on the same top
   holdings — active fees paid repeatedly for one exposure); allocation
   drift (does SIP flow still match the plan's target mix, or has it
   skewed, e.g. all new money into small-cap after a hot run); blended
   expense ratio vs. a passive equivalent.

4. **Verdict per SIP:** continue / continue-but-watch / pause / redirect
   to [bucket], each per the completeness checklist. Redirect targets are
   buckets; specific fund picks route through `mutual-fund-analysis`. Any
   change is the user's action in their broker's app.

5. **Present:** one summary table first (fund, SIP ₹, vs. benchmark, rank
   trend, flag, verdict), then detail only for non-continue verdicts,
   then set-level flags, then the disclosure block. Formal version:
   `reference/templates/sip-review.md` — worth saving for the next review.