Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add Sundeepg98/kite-mcp-server --skill tax-harvestgit clone --depth 1 https://github.com/Sundeepg98/kite-mcp-serverWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/sundeepg98/kite-mcp-server/tax-harvest)<a href="https://agentmods.dev/skills/sundeepg98/kite-mcp-server/tax-harvest"><img src="https://agentmods.dev/badge/skills/sundeepg98/kite-mcp-server/tax-harvest/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/sundeepg98/kite-mcp-server/tax-harvest"><img src="https://agentmods.dev/badge/skills/sundeepg98/kite-mcp-server/tax-harvest.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00109 | $0.01465 |
| Opus 5 | $0.00055 | $0.00732 |
| Sonnet 5 | $0.00022 | $0.00293 |
| Haiku 4.5 | $0.00011 | $0.00146 |
Grade A, and why
tax-harvest scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 12d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 116 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Tax-Loss Harvesting (India)
Indian equity capital gains tax is layered (STCG vs LTCG, different rates, ₹1.25L LTCG exemption per financial year). This skill helps the user identify positions sitting on losses and plan realizations without tripping over Indian-specific rules.
1. Pull the data
- Call
tax_loss_analysis— server-side analysis that buckets holdings into STCG/LTCG, flags loss candidates, and estimates potential tax savings at the current effective rates. - Call
get_holdingsto cross-check quantities and avg prices. - Call
get_tradeswith a date range to see realized gains already booked this financial year (if the user has been trading). Needed to know how much tax shield is actually useful. - Call
get_ltpon each candidate to price the exit.
2. Indian capital gains tax basics (use as facts)
- Financial year: April 1 to March 31. Harvesting must be settled before March 31 to count for the current FY.
- STCG (Short-Term Capital Gains): Equity held ≤12 months → taxed at 20% flat (revised from
15% — confirm current rate from
tax_loss_analysisresponse, which reflects latest rules). - LTCG (Long-Term Capital Gains): Equity held >12 months → ₹1.25 lakh annual exemption per individual, then 12.5% on the excess (revised from 10%).
- Losses: STCL can offset STCG or LTCG. LTCL can only offset LTCG. Carried forward up to 8 years if filed on time.
- No wash-sale rule in Indian tax law as of this writing — you can sell and re-buy the same day, and the loss is still recognized. BUT the re-buy resets the holding period to zero, so an LTCG position becomes STCG. Consider this carefully.
- STT, brokerage, exchange charges are deductible from capital gains (use
get_order_chargesto estimate these).
3. Format the harvest plan
## Tax-Loss Harvest Plan — FY <YYYY-YY>
### Current tax position
- Realized STCG this FY: <amt>
- Realized LTCG this FY: <amt> (₹1.25L exemption: <used | unused>)
- Realized STCL / LTCL this FY: <amt>
- Net taxable so far: <amt>
### Candidates for harvesting (unrealized losses)
| Symbol | Qty | Buy Date | Avg Cost | LTP | Unrealized Loss | Bucket |
|---|---|---|---|---|---|---|
| <sym> | <q> | <date> | <price> | <price> | <loss> | STCL | LTCL |
### Estimated tax shield
- STCL available to book: <amt> → offsets <amt> of STCG at 20% = ₹<tax savings>
- LTCL available to book: <amt> → offsets LTCG at 12.5% = ₹<tax savings>
### Proposed sequence
1. SELL <qty> <symbol> (books <loss amount> as <STCL|LTCL>)
2. ...
### Re-buy consideration
<If user wants to keep exposure:>
- You can re-buy the same symbol immediately (no wash-sale in India).
- IMPORTANT: Re-buying resets the holding period clock to zero.
- If you sold a position held 14 months (LTCG), re-buying means the new lot is STCG until it
crosses 12 months. This may cost more on future gains than the loss you're booking now.
- Alternative: buy a similar-exposure stock or sector ETF to preserve market exposure without
resetting the clock on the sold name.
### Charges to factor in
- STT on sell: ~0.1% (delivery)
- Brokerage: <user's plan — check with `get_order_charges`>
- Exchange charges, stamp duty, GST: small but add up
- Net benefit after charges: <amt>
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 12d ago First seen · 116 lines · 109 tokens per session scan A 9441781bc034
tax-harvest is a skill published in the GitHub repository Sundeepg98/kite-mcp-server (0 stars, last pushed 3mo ago), licensed MIT. It adds 109 tokens to every session and 1,465 once invoked, about $0.0005 per session on Opus 5. A static security scan graded it A with 0 findings. It comes from a forked repository.
Other skills, from other repositories
quant-mathematician
You are a world-class applied mathematician, statistical researcher, and quantitative strategy architect.
position-optimizer
You are a top 0.1% quantitative position-sizing and risk-optimization agent.
ai-trading-agent
You are an AI trading research assistant connected to Trader Dev MCP.
strategy-optimizer
You are a top 0.1% quantitative strategy optimization agent.
mean-reversion-engineer
You are a top 0.1% quantitative trading engineer and Pine Script developer.
ai-hedge-fund
You are the manager of an AI-powered hedge fund research desk.