Agent skill

Trade Pitch

by ginlix-ai in ginlix-ai/LangAlpha

Turn finished analysis on one named security into a position: variant view, falsifiable evidence, scenario tree, expression, risk and monitoring.

Apache-2.0Auto-check passedBusiness, Finance & HR

Install Trade Pitch

skills CLI
$ npx skills add ginlix-ai/LangAlpha --skill trade-pitch -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI
$ gh skill install ginlix-ai/LangAlpha trade-pitch --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Manual copy
$ git clone --depth 1 https://github.com/ginlix-ai/LangAlpha.git skills-src && mkdir -p .claude/skills && cp -r skills-src/plugins/langalpha_research/skills/trade-pitch .claude/skills/trade-pitch && rm -rf skills-src

Use ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.

Claude Code skills documentation · loads skills from .claude/skills/

Facts

Skill name
trade-pitch
GitHub stars
1.8k
Token cost
~3.5k tokens
SKILL.md length
2,109 words
Files
2 (incl. references)
Skills in repo
37
Repo updated
First seen
Licence
Apache-2.0

At a glance

Turn finished analysis on one named security into a position: variant view, falsifiable evidence, scenario tree, expression, risk and monitoring.

  • Works in 7 steps: Intake, and the posture the inputs support → The variant view → Evidence as falsifiable claims → …
  • Pitch this name
  • SKILL.md covers Step 1: Intake, and the…, Step 2: The variant view, Step 3: Evidence as… and Step 4: Scenario tree, plus 3 more sections
  • Instructions only: no scripts, shell commands, URLs or credentials in SKILL.md

What it does

Trade Pitch is an agent skill from ginlix-ai/LangAlpha. Turn finished analysis on one named security into a position: variant view, falsifiable evidence, scenario tree, expression, risk and monitoring. Triggers on pitch this name, should I buy, a long idea or short idea on a named stock, position, trade recommendation, risk reward.

Its SKILL.md is about 3.5k tokens, which your agent loads only when the skill is triggered. The skill folder holds 2 other files, including reference files (for example `references/expression.md`).

It sits in Business, Finance & HR. The repository describes itself as: Claude Code for Financial Market. The licence is Apache-2.0.

When your agent uses it

  • Pitch this name
  • Short idea on a named stock
  • Trade recommendation

Example prompts

  • “/trade-pitch”

Workflow steps

7 steps, taken from the step headings in SKILL.md.

  1. Intake, and the posture the inputs support
  2. The variant view
  3. Evidence as falsifiable claims
  4. Scenario tree
  5. Trade expression
  6. Monitoring
  7. Output

What it can do on your machine

Read from SKILL.md and the folder at commit 2855e43. It shows what the files ask for, not the result of running them.

  • Tool permissions

    Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

    From allowed-tools in the SKILL.md frontmatter.

  • Runs code

    No scripts in the folder and no shell commands in SKILL.md.

    From the folder's file list and the shell code blocks in SKILL.md.

  • Network

    No URLs in SKILL.md.

    From URLs in SKILL.md, links to its own repository left out.

  • Credentials

    Names no API keys, tokens, secrets or passwords.

    From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Trade Pitch loads about 3.5k tokens when it runs, and up to ~6k if it reads all its reference files. Until then it costs about 72 tokens; SKILL.md has 2,109 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~72
When it runs · the whole SKILL.md, loaded when a task matches
~3.5k
With references · SKILL.md plus every file in references/, read only if the agent opens them
~6k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Auto-check passed

The automated check found no risky patterns in SKILL.md.

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.

SKILL.md

The full file from ginlix-ai/LangAlpha at commit 2855e43, republished under its Apache-2.0 licence (© ginlix-ai). 2,109 words, ~3,527 tokens.

Download SKILL.mdSave it as .claude/skills/trade-pitch/SKILL.md (or your agent's skills folder). This skill also uses 1 other file; get the full folder from GitHub.
name
trade-pitch
description
Turn finished analysis on one named security into a position: variant view, falsifiable evidence, scenario tree, expression, risk and monitoring. Triggers on pitch this name, should I buy, a long idea or short idea on a named stock, position, trade recommendation, risk reward.

Trade Pitch

The terminus of the research. A finished analysis becomes a position: a direction, a variant view, the evidence that carries it, a scenario tree, an expression, a level that says we were wrong, and a plan for watching it. The pitch is research output for a professional reader, and its verbs describe what the analysis supports; personalised advice and execution instructions sit outside it, which is why the vocabulary ends at re-underwrite rather than at a size, and why size appears only as a fraction of a stated loss budget, per the registered extension in .agents/skills/research-conventions/references/judgment.md.

Every figure in a pitch is load-bearing, which makes this the deliverable where unsupported precision does the most damage. Evidence labels, source tiers, staleness, the readiness posture and the intake limits: .agents/skills/research-conventions/SKILL.md, read before the first figure enters the pitch.

Instrument mechanics for anything other than long common stock, the short-side gate, and the worked reconciliation of two loss limits: .agents/skills/trade-pitch/references/expression.md.

Step 1: Intake, and the posture the inputs support

Read the investor-mandate memory first per .agents/skills/research-conventions/references/intake.md, then ask only about the forks it leaves open. Four things are fixed before any work starts:

SlotTaken fromDefault when unstated
Directionthe requestthe direction the source analysis already leans, stated as an assumption
Horizonmandate memory, then the catalyst that closes the gaptwo to eight quarters
The analysis it rests ona named artifact: a dcf-model workbook, a comps-analysis, an earnings-analysis print, an initiating-coverage initiation, or the user's own stated viewnone; see below
Mandate slotsmandate memory: vehicle, benchmark, risk posture, scope, instruments alloweddisclosed in the delivery message

The pitch rests on something named. Record the artifact, its as-of, and which of its outputs the pitch uses (the fair-value range, the peer multiple, the revision path). Where no finished analysis exists, say so in one sentence, name which skill would produce it, and either build that first or deliver a screen-grade pitch that says the underwriting is still owed.

Posture is stated once near the top of the artifact, read from the ladder in .agents/skills/research-conventions/SKILL.md against the input state. For a pitch the inputs that count include a live price, a valuation anchor, a dated catalyst, an established borrow on a short leg, and the sizing inputs in .agents/skills/trade-pitch/references/expression.md.

Done when direction, horizon, the source artifact with its as-of, and every mandate slot carry a value or a disclosed default, and the posture is written with the specific input that set it.

Step 2: The variant view

Four fields, and a pitch without all four is a summary of someone else's work:

  1. What the market believes. At the current price, in the units of the driver the position rests on: the growth rate the multiple requires, the margin the price embeds, the estimate path consensus is carrying. Ground it in the affected names and their revisions, never in an index move.
  2. What we believe. The same driver, our number, sourced to the analysis from step 1.
  3. Why the gap exists. The mechanism that lets it persist: a disclosure lag, a KPI the street reads on the wrong basis, a misclassified business, a forced seller, a horizon mismatch, an event nobody has dated. A gap with no mechanism is usually an error in our own work.
  4. What closes it. The dated event or the evidence arrival that converts belief into price, marked scheduled, likely or speculative.

The seven questions this section has to satisfy, the rule that price is belief rather than truth, and the requirement to answer the strongest opposing case: .agents/skills/research-conventions/references/judgment.md, read before writing this step.

Done when all four fields are written, the market's belief is stated in the driver's own units and sourced to the affected names, the gap names a mechanism, and the closer carries a date or a window and its confidence.

Step 3: Evidence as falsifiable claims

Convert the argument into two to five claims, in the table of .agents/skills/research-conventions/references/judgment.md, with one column added: the evidence label from .agents/skills/research-conventions/references/evidence.md on each claim's evidence. The working table carries the research label; the delivered document prints the reader-facing word for it.

ClaimEvidence todayLabelImplication if trueTest metricFalsifierTime to knowable

A needs-source label is drafting state: before delivery the figure is sourced, or the claim takes the unsupported-claim exit in evidence.md.

Done when the table holds between two and five claims, each carries all six fields plus one evidence label, no needs-source label survives, and removing any one claim would change the recommendation.

Step 4: Scenario tree

Three cases, each built from a driver rather than from a haircut to the base target:

CaseDriver and levelArithmetic to priceReturn from spotProbabilityProvenance

Each case moves one named driver to a stated level and carries the arithmetic through to a price, so a reader can move the driver and watch the case move. Provenance is the model, a source, the user, our judgement or illustration, with an as-of. Returns are discounted or annualised against a stated hurdle per the discounting rule in judgment.md, because a multi-year target price is not a return.

Probabilities. A complete set is mutually exclusive, covers the space and sums to one; the rule and what to do otherwise are in judgment.md. When the probabilities are assumed rather than underwritten, the table is labelled an illustrative skew, the weighted figure is withheld, and the posture is re-read from the ladder in .agents/skills/research-conventions/SKILL.md against that input state.

Bear-case survivability. A lower price is not a downside case. State what breaks, through which line item, to what number, and then whether the company survives it: liquidity through the trough, covenant headroom against the stressed metric, and the maturity wall inside the horizon. A company that cannot fund its own bear case has a different downside from one that can, and the difference is the whole risk.

The exit in the bear case uses stressed liquidity, not normal. Normal participation and stressed participation are separate lines, never one blended figure, and where the case is a gap risk the exit price is the post-move price rather than the level we would have liked. Mechanics and the arithmetic: .agents/skills/trade-pitch/references/expression.md.

Done when each case names its driver, its level and the arithmetic to a price; the bear case states liquidity, covenants and maturities through the trough; the probability set sums to one or the table is labelled an illustrative skew with the posture re-read against that state; and the bear-case exit carries a stressed participation figure beside the normal one.

Show full SKILL.md (1,014 more words)Show less

Step 5: Trade expression

FieldCarries
Instrumentcommon, an options structure, or a pair, with the basis risk it leaves
Entrya rule rather than a quote ("at or below 48"), with the reference price and its as-of beside it, and the tranches where it is staged
Targetfrom the valuation anchor in step 1, as a range rather than a point, discounted or annualised against a stated hurdle, saying which end the position underwrites
Exitthree levels. A stop is a price rule, set where the price itself is evidence, its distance stated in units of the name's own volatility. A re-underwrite level is a claim rule naming the input that broke, taken from the falsifier column. A time stop is the horizon date at which the absence of the catalyst is itself the answer
Sizea fraction of a loss budget, per the rule below
Binding constraintwhich lens set the size: valuation, liquidity, mandate, conviction or the loss limit

Size against a loss budget, not a book. We hold no positions, no net asset value and no mandate limits, so a percentage of a book would be a number we invented. State size as what it risks against a budget the reader brings: "at the stop this position spends a third of a 150bp loss budget" travels to any book size, and names the binding constraint that set it per judgment.md.

The two loss limits are different constraints. A limit stated inside a stress scenario ("lose no more than this by the time the bear case has played out") and an absolute cap that must hold beyond it ("lose no more than this in any state") give different sizes and sometimes different answers about whether the position exists at all. Resolve which one the mandate means before recommending anything; where the mandate does not settle it, show both branches rather than picking one. The worked example is in .agents/skills/trade-pitch/references/expression.md.

The short-side gate. A short pitch carries a gate block with six rows, each marked cleared, not cleared, missing, n/a or illustrative: catalyst, valuation anchor, liquidity, borrow and carry, option cost, hedge ratio. Any applicable row below cleared is named in the recommendation itself rather than in a footnote, and an n/a row is not a gap. An unhedged short cannot satisfy an absolute cap, because there is no price at which the loss stops, so under that reading the answer is a priced defined-loss structure or no position. What clears each row: .agents/skills/trade-pitch/references/expression.md.

Every hedge carries the scenario in which it fails, per judgment.md, and two rules sit beside it: a hedge that removes the reason the position exists is rejected rather than priced, and the alternative of taking no hedge and sizing down is always stated so the reader can compare them.

Done when the six expression fields are filled, size is stated against a loss budget with its binding constraint named, the two loss limits are resolved or both branches are shown, a short carries all six gate rows with one of the five verdicts each, and every proposed hedge carries a failure scenario beside the size-down alternative.

Step 6: Monitoring

TriggerMetricThresholdSourceDate or windowAction each side

judgment.md sets the bar for a monitored item; the pitch adds that every claim in step 3 is testable by at least one trigger, so a falsifier without a trigger is an intention rather than a plan.

What would change our mind is its own short block: the two or three observations that would retire the position, each with the date it becomes observable, drawn from the falsifier column rather than written fresh.

Hand the running ledger to .agents/skills/thesis-tracker/SKILL.md: the pitch's claims become its pillars, the step 3 falsifiers become its break thresholds, and the triggers here become its monitoring rows. The pitch is a dated document; the tracker is what keeps it honest afterwards.

Done when every trigger row carries its metric, threshold, source and date, with an action stated on each side of the threshold; every step 3 claim is covered by at least one trigger, and the hand-off names the tracker and the first review date.

Step 7: Output

Actionability leads, background follows. The order, top to bottom: the recommendation line (action verb, direction, instrument, entry, target, exit, posture and as-of), the variant view in one sentence, the missing-inputs block, the scenario table, the expression and its risk, the claim table, monitoring, then the analysis the pitch rests on. A reader who stops after the first screen has the position; a reader who continues gets the underwriting.

The action verb comes from the closed vocabulary in judgment.md and is available only once its inputs are in hand. Without a live price, without sizing inputs, and with no position to act on, the honest verbs are watchlist, wait for proof and re-underwrite, each naming the missing input in the same sentence and saying what would convert the idea into an action.

Reader-facing evidence vocabulary. The label taxonomy stays in the working research and the document prints the reader's word for each label, one per label, per Reader-facing labels in .agents/skills/research-conventions/references/evidence.md.

The missing-inputs block is one compact block near the top, not a table of blanks. Each line names what is absent, what would supply it, and what it would change about the recommendation. A section whose input is unavailable keeps its heading and carries its labelled gap, because a short pitch reads as a thin idea while a full pitch with three labelled gaps reads as the truth.

Deliver one to three pages as Word through .agents/skills/docx/SKILL.md, or a deck through .agents/skills/pptx/SKILL.md when the user asked for slides. Save to {task}/. Depth bands and what to cut first when the draft runs long: .agents/skills/research-conventions/references/depth.md.

Done when the first screen carries the action verb, the position, the posture and the variant view; every label the delivered document prints is the reader-facing word for its research label; the missing-inputs block names each absent input with what would supply it; and the file is written to the task directory.

© ginlix-ai, Apache-2.0. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

SKILL.md and 1 other file (references) in plugins/langalpha_research/skills/trade-pitch of ginlix-ai/LangAlpha.

  • SKILL.md
  • references/expression.md

Open the folder on GitHubat commit 2855e43

Compare with similar skills

Trade Pitch next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.

Trade Pitch compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Trade Pitch this skillginlix-ai/LangAlpha1.8k—~3.5kAutomated safety check: PassApache-2.0
Technical Analysttradermonty/claude-trading-skills3k4 repos~4.6kAutomated safety check: PassMIT
Theme Detectortradermonty/claude-trading-skills3k2 repos~4.9kAutomated safety check: PassMIT
Creating Financial ModelsChen-zexi/open-ptc-agent7293 repos~1.3kAutomated safety check: PassMIT
Stock APIzhangxiangliang/stock-api2k—~507Automated safety check: PassMIT
Itr Walakaranb192/itr-wala871—~3.6kAutomated safety check: PassMIT

Similar skills

  • Technical Analyst

    tradermonty/claude-trading-skills

    This skill should be used when analyzing weekly price charts for stocks, stock indices, cryptocurrencies, or forex pairs.

    3k GitHub starsUsed in 4 repos~4.6k tokens
    Business, Finance & HRAuto-check passed
  • Theme Detector

    tradermonty/claude-trading-skills

    Detect and analyze trending market themes across sectors. An agent skill from tradermonty/claude-trading-skills.

    3k GitHub starsUsed in 2 repos~4.9k tokens
    Business, Finance & HRAuto-check passed
  • Creating Financial Models

    Chen-zexi/open-ptc-agent

    This skill provides an advanced financial modeling suite with DCF analysis, sensitivity testing, Monte Carlo simulations, and scenario planning for investment decisions

    729 GitHub starsUsed in 3 repos~1.3k tokens
    Business, Finance & HRAuto-check passed
  • Stock API

    zhangxiangliang/stock-api

    Fetch real-time stock quotes, K-line (candlestick) history, and search symbols for China A-shares, Hong Kong, and US markets.

    2k GitHub stars~507 tokensUpdated yesterday
    Business, Finance & HRAuto-check passed
  • Itr Wala

    karanb192/itr-wala

    File Indian income tax returns (ITR) for FY 2025-26 / AY 2026-27.

    871 GitHub stars~3.6k tokensUpdated 7 days ago
    Business, Finance & HRAuto-check passed
  • Tushare Data

    zillionare/zillionare

    面向中文自然语言的 Tushare 数据研究技能。用于把“看看这只股票最近怎么样”“帮我查财报趋势”“最近哪个板块最强”“北向资金在买什么”“给我导出一份行情数据”这类请求,转成可执行的数据获取、清洗、对比、筛选、导出与简要分析流程。适用于 A 股、指数、ETF/基金、财务、估值、资金流、公告新闻、板块概念与宏观数据等研究场景。

    322 GitHub starsUsed in 2 repos~2.3k tokens
    Business, Finance & HRAuto-check passed

More from ginlix-ai/LangAlpha

All 37 skills in this repo
  • Investment Deck Check

    ginlix-ai/LangAlpha

    Quality-checks an investment deck in .pptx form before it goes out: number consistency, chart and narrative alignment, source coverage, language and a circulation verdict.

    1.8k GitHub stars~3.7k tokensUpdated today
    Auto-check passed
  • Equity Initiation Report

    ginlix-ai/LangAlpha

    Produces a first-time equity research initiation report in five tasks: company research, financial model, valuation, charts and a DOCX report.

    1.8k GitHub stars~3.8k tokensUpdated today
    Auto-check passed
  • Builds or repairs an integrated income statement, balance sheet and cash flow model in Excel with live formulas, supporting schedules, scenarios and a Checks sheet.

    1.8k GitHub stars~5.4k tokensUpdated today
    Auto-check passed
  • Financial Model Checker

    ginlix-ai/LangAlpha

    Audits an existing Excel financial model without editing it, checking structure, formulas, integrity identities and source tie-out, and ends in a prioritized issue log.

    1.8k GitHub stars~4.2k tokensUpdated today
    Auto-check passed
  • DCF Model Builder

    ginlix-ai/LangAlpha

    Builds a live Excel DCF valuation workbook with free cash flow projections, WACC, terminal value, three scenarios, sensitivity grids and a reverse DCF.

    1.8k GitHub stars~7.7k tokensUpdated today
    Auto-check passed
  • Builds Word files with python-docx, edits existing ones in place with tracked changes and comments, then renders and validates the result.

    1.8k GitHub stars~4.8k tokensUpdated today
    Auto-check passed

Questions about Trade Pitch

What does Trade Pitch do?

Turn finished analysis on one named security into a position: variant view, falsifiable evidence, scenario tree, expression, risk and monitoring. Trade Pitch is an agent skill from ginlix-ai/LangAlpha. Turn finished analysis on one named security into a position: variant view, falsifiable evidence, scenario tree, expression, risk and monitoring.

When should I use Trade Pitch?

Trade Pitch fits situations like: pitch this name; short idea on a named stock; trade recommendation.

How do I install Trade Pitch in Claude Code?

Run `npx skills add ginlix-ai/LangAlpha --skill trade-pitch -a claude-code`. Or copy the skill folder (plugins/langalpha_research/skills/trade-pitch in ginlix-ai/LangAlpha) into .claude/skills/trade-pitch in your project. Claude Code loads it when a task matches its description.

How do I install Trade Pitch in Codex?

Run `npx skills add ginlix-ai/LangAlpha --skill trade-pitch -a codex`. Or copy the skill folder (plugins/langalpha_research/skills/trade-pitch in ginlix-ai/LangAlpha) into .agents/skills/trade-pitch in your project. Codex loads it when a task matches its description.

Can I use Trade Pitch in Cursor, Gemini CLI or GitHub Copilot?

Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add ginlix-ai/LangAlpha --skill trade-pitch -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/trade-pitch, .gemini/skills/trade-pitch, .github/skills/trade-pitch and .opencode/skills/trade-pitch in your project.

What does Trade Pitch need to run?

SKILL.md names no scripts, command-line tools or credentials: Trade Pitch is instructions for the agent only.

Does Trade Pitch access the network?

SKILL.md contains no URLs. Any network use would come from the scripts or tools the agent runs. This is read from the text; nothing was executed.

Is Trade Pitch safe to install?

Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. Review the folder before installing.

What licence does Trade Pitch use?

Trade Pitch is published under the Apache-2.0 licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does Trade Pitch use?

About 3.5k tokens (SKILL.md is roughly 14k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full. Its references folder adds about 2.4k tokens, read only when the agent opens those files.

What are the alternatives to Trade Pitch?

Skills that share tags, products or a category with Trade Pitch: Technical Analyst (tradermonty/claude-trading-skills, 3k stars), Theme Detector (tradermonty/claude-trading-skills, 3k stars), Creating Financial Models (Chen-zexi/open-ptc-agent, 729 stars) and Stock API (zhangxiangliang/stock-api, 2k stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Trade Pitch?

ginlix-ai (a GitHub organization) maintains it in ginlix-ai/LangAlpha, which has 1,811 GitHub stars. The repository holds 37 skills in this directory. The repository was last updated on October 10, 2026.

Source: ginlix-ai/LangAlpha on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.