Adapted from the DGF Agent Basics series — consumer-first guides for working with a personal AI agent. Pick only what you need.
> Pick only what you need. This skill is individual. This is one person's thoughts, offered as a pre-configured help, not a must. Take only the value you need from it, prune it to what fits you, and leave the rest. You do not have to use what the author thinks, and you do not have to take any other skill with it.
Summary: Research stocks and other investments step by step, with a bull versus bear case, risk, and watch triggers. No guarantee, research only, your agent never trades. Optional.
Use this skill when the user wants to find, research, or compare a stock or other investment. It is research and education only. It is not financial advice, and it never makes a trade for the user.
This skill stands alone. An agent that knows nothing else can follow it from the startup sequence below, then the investing method. This skill works on its own, with no other skill needed, and no knowledge of any other skill assumed.
Startup and how to use
Start by getting your own personal agent. Get your own personal agent and set it up. Bring your own accounts and logins. Use nothing that belongs to anyone else.
Tell your agent what you care about in plain words. Name your business, your work, your family, your hobbies, or any topic you want it to follow. Talking is the setup. You do not have to set up a system first.
Ask for the plan first, before any work runs. Say the task, then ask for the exact steps it will take, what it needs from you, and anything it cannot do. Wait for you to approve or correct the plan. Review the plan like a checklist and fix a wrong step or add a missing step before work starts.
Check that your agent understood you, before it runs. Ask it to show back what it thinks you mean and what output it will give you. Correct its repeat back right there when it does not match what you pictured.
Work in small single steps. Count the steps first. Give one step, check it, then give the next step. A ten step run is a huge run for an agent, so break it down and keep each step to exactly one thing.
Show it once, then save it by name. This is how you create a new task or a new skill. Take over the browser and do the task once yourself while your agent pauses. You drive alone and your agent does not watch your clicks. There is no record button to press. After you hand control back, tell your agent the steps in chat and tell it to save them by name, such as save this as Fun Things. Later, say pull up Fun Things and it picks up where you left off.
Correct it in small plain words when it does it wrong. Do not start over. Say what to change and say to remember that correction. That preference sticks and it gets better for you specifically over time.
Let it keep learning your interests every day. Your Feed looks like a newspaper with articles and topics related to what you told it you care about, updated daily. Your Ideas look like a lightbulb with suggestions for things you might like. It is not mind reading. It learns from what you actually talk about.
Use easy first and deep only when you want it. Never require a deep setup to use a basic task. Keep the wording in plain benefits and do not lead with virtual machine or VM. Say it has its own browser and workspace built in, you do not have to set anything up, it keeps your logins safe, and it saves and remembers your stuff.
No guarantee, stated up front and at every recommendation
Say this in plain words before any stock research, and repeat it with any recommendation. Your agent cannot guarantee anything. No research, no agent, no backtest, and no past return guarantees a future return. A stock can fall, a company can fail, and money invested can be lost.
Say that this is research and education, not financial advice. The user makes their own decision, in their own broker account, with money they can afford to lose. Never tell a user to invest money they need for rent, bills, or debt.
Never present a past return, a backtest, or a model score as a promise. State the date range, the source, and what could make that result not repeat.
Never execute a trade, move money, or connect to a broker account without the user's explicit go ahead for that exact trade, and even then the user reviews and places it themselves. Research ends at the report. The trade is theirs.
Setup questions, before any stock is named
Ask what the money is for, and when they would need it back. A short horizon and a long horizon are different jobs.
Ask how much they can afford to lose without harming their life, and how they react when a holding falls. Record risk tolerance in their own words, not as a score your agent invented.
Ask what they already hold, what broker or app they use, and whether they want to name an app or be steered to a connected route. If they name an app, use that app. If they have no app in mind, offer the routes your agent can check, such as a finance quote search and a broker they already connect, and check that connection before promising anything from it.
Ask what kind of company they are interested in, and what they will not invest in. Do not pick a sector for them.
Ask what they want the report to cover, and what decision it should help them make. A report with no decision question drifts.
Sequence for finding a stock
Run the sequence in order. Do not skip to a buy or sell call after one good headline.
Step 1, screen for candidates. List companies that fit the user's goal, horizon, and interests, with ticker, exchange, and what the company actually does in one plain sentence. State where each candidate came from, such as a sector list, a news story, or the user's own idea. Drop any candidate whose business the user does not understand after that sentence.
Step 2, fundamentals. For each remaining candidate, check what the company earns, whether it makes money, how much debt it carries, how fast it is growing, and what could break the business. Use filed financials where available, such as SEC filings for US companies, and state the filing date. Flag any figure that is stale, restated, or estimated, and say so.
Step 3, technical picture. Check the price trend, trading volume, and simple indicators such as moving averages, MACD, and RSI, explained in plain words. State the price and the date it was checked. A chart pattern is a description of the past, not a prediction, and must be labelled that way.
Step 4, news and macro. Check recent company news, earnings, and wider market conditions that could move the stock. Separate a fact from a rumour, and separate company news from market noise.
Step 5, sentiment, with caution. Check what people are saying in news and social discussion, but treat hype as a warning, not as proof. A crowded trade is a risk to name, not a reason to follow it.
Step 6, bull versus bear debate. Write the strongest case for the stock and the strongest case against it, each with its evidence. If the bear case cannot be answered with evidence, say so plainly. The user sees both sides before any decision.
Step 7, risk and portfolio fit. Check how this holding fits what the user already owns, how volatile it is, how easily it can be sold, and how large a loss would be if it fell. Suggest a position size as a research range tied to their risk tolerance, never as an order to place.
Step 8, decision and watch triggers. State what would make the case stronger, what would make it weaker, what price or news event would change the view, and when to review it again. A report without a review date goes stale silently.
Diverse, not concrete: let the user prune
Do not make this skill concrete, narrow, or one size fits all. Present a diverse set of options first, across price, style, risk, format, and approach, so different people can see themselves in it.
Let the user prune. Ask what to cut, what to keep, and what to change, and then narrow to what they kept. Pruning is the user's job, not your agent's guess. Do not prune for them before they have seen the options.
Do not lock a single choice early. Offer at least two or three meaningfully different options at each decision point, with the trade for each, unless the user already locked that choice themselves.
Keep options that do not fit one person, because they may fit another person using the same skill. A skill that only works for one kind of person is too concrete to reuse.
When the user prunes, save the pruned version by name when they want to repeat it, so their version starts from what they kept, not from the full diverse set again.
Present the result
Lead with the no guarantee line, then the trade off in plain words. Name the top candidate, its bull case, its bear case, and its biggest risk, in that order.
State every price with its date and source. State when a price, a filing, or a news item is stale, and do not present an old quote as current.
List sources with dates for every decisive claim. A claim with no source is labelled as opinion, not as fact.
End with the user's decision, not your agent's order. Ask whether they want to research further, watch it, or drop it. Stop after a denial, and do not re-pitch a rejected stock in new words.
Save the report by name when the user wants to track it, so the next review starts from the same report and its watch triggers.
Available connectors
Use your agent's connector and skill search, if it has one, to find other connectors and related skills.
If nothing here fits, look on the web for a maintained API or MCP server and set it up with your agent's custom connector setup, if it has one, or build a custom connector.
Websites
Use the finance search route for current quotes, and verify a decisive price on the broker or exchange page before presenting it as current. Search results alone do not confirm a live price.