What it is
The Trade Agent is a signal source you add to a bot, like TradingView or Telegram. It watches the bot’s pair and finds its own entries, or it judges the signals of one other source on the bot, and it decides an entry with its reason. What happens next depends on the tier you set: at Suggest nothing is placed until you tap Place it; at Enter it places the entry itself and tells you after. Its decisions are written down, so you can always see what it saw, what it decided, and why. There are two kinds of agent, and they work differently:The agent sends the entry and nothing else. Your bot’s own settings decide the rest, as they do for every source: Position size, Martingale, the Exit rule, the Filter, the Schedule and the Limits. Nothing you set on the agent can overrule the bot.
Adding it to a bot
- In the builder, press + under Signal sources and choose Trade Agent.
- Set it up in the panel (below) and press Save. Closing the panel with Done also saves any changes.
- Activate the bot. A sheet tells you which account it runs on, when the first look is, and what it can spend.
The panel
- Today — for Pip finding its own entries: “Free. Pip scans 37 setups. No agent credits are used.”, with the setup it is waiting for. For an agent that uses a model: the looks it took today and the agent credits you have left; at 20% left the line turns into a warning with Add credits.
- 1 · Agent — Pip, the built-in agent that works on every market (free, no agent credits), or an agent you built in Agents.
- 2 · Autonomy — two tiers. Suggest: the agent proposes, you approve each trade. Enter: it places the entry without a tap and you are told after each one; a gate can still refuse it. There is no tier where the agent runs the exits: every stop, target and trail is your bot’s Exit rule. Prove Enter on a demo account before a live one.
- 3 · It judges — where its signals come from. Pick one source on this bot (TradingView, Telegram, Discord and so on) and every signal from that source is held for the agent, which takes it or skips it with a reason; the bot’s other sources are not judged and trade as usual. Or pick Nothing: it finds its own entries and it watches the bot’s chart. One agent on a bot can judge.
- 4 · Setup (Pip finding its own entries) — one row, Pip picks the setup: the chart and the win rate it scans with. Described under How Pip finds entries.
- 4 · Feeds (an agent that uses a model) — what else the agent is given. One row today, News (Standard and up): described under “News for the agent” below.
- Limits · the bot’s, for every source — a read-only list of what will hold every order: Position size, Exit rule, Martingale, Filters, Schedule and Daily caps. Change them on the bot’s own panels.
- Model (an agent that uses a model) — the deciding model, and what a look costs: Sonnet is the default at 1 agent credit a look; Opus is the biggest model at 5 credits a look. Pip finding its own entries has no Model step: it asks no model.
- Tell me about — which events you are told about, and a Telegram row: link the TraderFrame bot once and proposals arrive on your phone with Place it / Skip, entries and daily reports as a line. Answering there is the same as answering in the app; a tap in both places cannot place twice.
- Look now (an agent that uses a model) — one look right away, so you can watch it think.
How Pip finds entries
Pip, finding its own entries, is a scanner. It asks no AI model and uses no agent credits.1
Scan
Pip replays its 37 setups over the last 24 hours on the chart you picked. A setup with fewer than five finished trades in that time is judged on its last 48 hours instead. Each setup is tried both ways, followed (trade with its signal) and faded (trade against it), so 74 ways in all. The setups are plain ones: breakouts, streaks, RSI and Bollinger extremes, average crosses, candle shapes, sweeps, session opens, the day’s average price, volume and momentum.
2
Pick
Each past signal is settled with your bot’s own Exit rule: it counts as won when price reached your take profit, or your first trailing step if you set a trail, before your stop loss. Every replayed trade also pays what a real one would on your own broker: its spread and, on a crypto exchange, your account’s taker fee on both sides. A setup needs at least five finished trades to count, and it must have come out ahead after those costs. Among those at or above your win rate, Pip takes the one with the best record.
3
Wait
Pip watches every close of your chart for that setup’s next signal.
4
Send
When the setup fires, Pip sends buy or sell to the bot: a proposal at Suggest, an entry at Enter. The bot sets the size and the exits. One trade at a time.
5
Scan again
After that trade closes, Pip scans again, so it does not stay on one setup.
- No setup reached your win rate. It sends nothing and scans at every candle close until one does.
- The picked setup gave no signal for 24 hours. It drops it and scans again.
- You changed the chart or the win rate.
The Setup view
Open step 4 · Setup in the panel.
Below them the view shows What to expect, worked out from this market’s own candles with your bot’s exits: trades a month, trades won, and what your Exit rule needs to break even. Under it, Costs counted shows the spread and fees it read from your broker and what each trade pays for the round trip. On MetaTrader only the spread is counted, because the terminal does not send your broker’s commission. It warns you when the replay won less than the exits need. Your plan compares the trades a month with your plan’s monthly trades.
Before Pip trades a live account
On a live (real-money) account, Pip asks once before it places its first order. The pop-up shows the account, the market and chart, and the same estimate and costs as the Setup view.- Yes, let Pip trade live: Pip places orders from then on.
- Not yet, show me its picks: Pip keeps scanning and shows each pick as a card for you to tap. It places nothing by itself, and the agent’s Today card reads Waiting for your OK with a button to let it trade.
What Pip writes on Agent decisions
A row appears only when something changes, not at every candle:
If an exchange refuses a candle request for rate limiting, Pip asks again every 30 seconds for about two minutes before it writes a row. Nothing is traded on a look that could not run, and the next candle close tries again.
What is no longer on the agent
The agent used to carry limits of its own. They are gone, and anything saved for them is ignored, not carried over:
Style, Look every and the timeframes are no longer choices on the bot either.
What it sees
This section and the next describe a look by an agent that uses a model: Pip judging another source’s signals, or a classic agent taught in words. Pip finding its own entries reads only the candles of the chart you picked, and an agent you built runs its own script. Every look is one snapshot, built from public data and the bot’s own record. Nothing in it comes from your exchange account.- Candles — the last 100 closed bars on the entry timeframe and 60 on each higher one: open, high, low, close, in UTC. The bar still forming is left out.
- Read from the candles — ATR(14), session high and low and where price sits between them, structure (higher lows, lower highs, range, mixed), the last two swings, bars since the session extreme.
- Volume — on crypto, the last bar and the last five bars as a share of the typical bar (the median of the 20 before). Forex and Deriv feeds carry no volume.
- The clock — the session it is in, and the next red news release and how far away it is. This is only for its reasons: the agent has no news pause of its own. Whether the bot trades around a release is decided by your bot’s Filter.
- The bot — open position or not, trades today, your Trades a day if you set one, ladder step, day P&L, last loss.
- Its memory — its own last 20 decisions on this bot with their outcomes.
- Spread and costs — bid and ask, as a price and as a share of price, with 24-hour turnover, and the round-trip cost (spread plus fees on crypto). The agent does not refuse a trade for its cost or for a stop that sits close to the spread: your Exit rule sets the target, so on a market with a wide spread (a thin demo market, for example) check that your target clears it.
- Funding — on perpetuals, the current rate and which side pays. Beyond ±0.05% per interval the paying side is crowded: an entry on that side is a skip unless it is a pullback in a higher-timeframe trend.
- Depth — the resting orders within half a percent of price, folded into a bid share and the nearest large wall on each side. A target is placed before a wall, never past it. The book alone is never a reason to enter.
What it reads on every close
Before any model is called, the engine reads the chart the way a discretionary day trader would, and every number goes into the look:- Levels — support and resistance from clusters of swing highs and lows, with how many times each held. Liquidity — equal highs or lows where stops rest, and sweeps: a wick through a level that closed back inside.
- Gaps — fair value gaps (three-candle imbalances not yet filled) and opening gaps away from the prior close.
- The candle — pin bar, engulfing, inside bar, strong close. The regime — EMA20/50 alignment, volatility against its own average, the last five swings as HH/HL/LH/LL, the bias, the last break of structure and any change of character.
- Context — a wider timeframe behind its two (4h behind 5m/1h) for the bias and the levels that matter, plus the prior day’s high, low and close.
Each playbook carries a quality score and its evidence in words. A close with no playbook is a free engine skip that says what it saw. A found setup goes to the model, which takes the best one when it is sound — quality 0.55 or more, the stop beyond the structure — or names the flaw. “Mid-range” on its own is no longer a reason to skip. On the Decisions tab the playbook shows as a chip on the row, and Open the look lists everything that was on the table.
The daily report prices looks the way you are charged. “The agent’s day” states the looks as a count times the look rate on your plan, the per-look price of the pack at your plan’s scale (15¢ a look up to 100 looks a month, 4.9¢ up to 1,000, 4.3¢ up to 3,000, 3.5¢ beyond). It never shows what the model itself cost.
Quiet rows stay out of the list. While the bot’s trade is open the agent keeps looking on its clock but writes nothing; the open trade is on the Trades tab. A sleep that lasts until something changes (your credits are used, the bot’s Trades a day is reached) appears once, not every five minutes.
Every look can be drawn. Open a look and the chart at the top shows the candles the agent was given, the levels that have held, the pools of stops, the sweep and the open gaps, with the trade drawn over them: Entry, Stop, each trailing step (where profit locks), and the Take profit. Numbered pins mark the playbook’s level, the sweep, the walls on the way, and the prices the reason names, with the text under the chart. On a skip the chart draws the best playbook that was on the table, labelled as not taken.
Your Exit rule sets the exits. When the bot executes at market, every agent entry carries the stop, target and trail from the bot’s own Exit rule, and the agent picks entries that fit them — with a trail on, it prefers moves with the momentum to run through the first step: points on MT5 (gold has two digits, so TP 1000 points is $10.00 and SL 500 points is $5.00), ROI percent of margin on crypto. The agent does not invent its own targets there; it judges whether price can reach the point where your profit locks before your stop, and it never skips because the target is small or far. That point is the first trailing step when your rule has a trail, else the take profit. So a rule of TP 2000, SL 300 and trailing at 450 points on gold is read as 300 of risk against 450 of reach, with the 2000 left as the big move the trail is allowed to catch. A level that has held before and sits on the way is named to the agent as a wall to weigh, not an automatic skip. A bot set to Limit execution works the same way: the agent’s entry is the limit price, and the stop and target are your Exit rule’s. To scalp gold for 1000 points with a 500-point stop, set the Exit rule to TP 1000 and SL 500 points and keep execution at market.
The agent also keeps a record of how each playbook has done on your bot over the last 30 days. The record is shown to the agent for information; it does not hold back a playbook because of its past results on your bot.
News for the agent (Standard and up). Step 4 of the agent’s panel, Feeds, has one row: News. Open it and switch it On, then save. With every look the agent is then told the last high-impact economic release for the bot’s currencies from the past four hours: what it was, how long ago, whether it came out above or below its forecast, and whether the surprise was big. The opened look on the Agent decisions tab shows the same under News, for example “25 min after USD Inflation Rate MoM: below forecast, a big surprise”. It is context only: it never says which way to trade, places nothing, and changes no size or limit. It is read from the economic calendar and costs no agent credits. Headlines and social posts are not read. Whether the bot trades around a release is still your bot’s Filter.
When it seems to never trade
Pip finding its own entries trades only when its picked setup fires. A row reading No pick yet means no setup reached your win rate and came out ahead after costs in the last 24 to 48 hours: a lower Won at least or a different Chart in the Setup view gives more picks, though not a better record. A row reading Picked means Pip is waiting for that setup’s signal. An agent that uses a model: most looks end in a skip; that is the design. The strip on the Agent decisions tab shows the week in numbers — looks, candidates, entries — and, when there were looks but no entries, a Why quiet line with the latest reason. A run of engine skips reading “No playbook: …” is a tape with no level, no sweep and no rejection candle to trade, not a fault. If you want more setups, check that the bot’s Schedule covers the hours the instrument moves.Looks and agent credits
A look is one model call by your agent. It is paid in agent credits, the agent’s unit the way a trade is the bot’s: 1 credit a look on Sonnet (the default model), 5 credits a look on Opus.- Pip finding its own entries is free. Its scan asks no model, so it uses no agent credits, however often it scans. Agent credits are used when Pip judges another source’s signals, when Pip works on an agent you built (see What costs agent credits), and by a classic agent taught in words.
- The checks the engine settles on its own — outside the bot’s Schedule, a position already open, Trades a day reached — are not charged. They are free and still logged, once per streak.
- A look that becomes a trade you place counts its credits and one trade.
- On a 7-day trial the agent has no credits: it keeps its free engine checks but makes no model calls, and credit packs can’t be bought. The plan’s credits and models start when the trial converts; the panel says “no agent credits on a free trial” and names the date. To have the agent from day one, choose Pay now at checkout instead of the trial.
- The Trade Agent is on Basic and up; the Free plan has none. Paid plans include agent credits a month: Basic 50 · Standard 150 · Pro 400 · Ultimate 1,000. On Opus that is 10, 30, 80 and 200 looks. Plan credits reset with the billing period.
- Credit packs top a plan up: 100, 1,000, 3,000 or 10,000 credits bought outright on the pricing page (Add Trade Agent credits). Pack credits never expire and are used only after the month’s allowance is gone. They stay on your account if your plan ends or changes; only the plan’s monthly credits end with the plan. Add credits appears on the agent panel when credits run low and in the pop-up when they run out.
- A look is refused before it starts when fewer credits remain than the bot’s model needs. An Opus bot with 3 credits left sleeps until the next period or a pack; switching it to Sonnet lets it use the rest.
- There is no daily cap on looks. The agent keeps watching while the bot runs, until your agent credits run out.
- At 20% left you get one warning: on the bell, on Telegram if you linked it, and in the agent panel, each with Add credits. Pack credits count as credits left.
- When they run out the agent stops watching and your bot keeps trading its other sources. You are told once, and a pop-up in the bot builder offers a credit pack or the next plan up. The agent starts again by itself the day your plan’s credits renew, or as soon as you add credits.
- Where to see them: the sidebar plan card, Subscription & billing, the agent panel (which also shows what a look costs on the chosen model), and the Agent decisions tab, where each row says what it charged.
Proposals
When the agent decides to enter, you get a card in your notifications with Skip and Place it, the reason, and the entry, stop and target. It expires after 5 minutes; a stale entry on a fast market is a different trade. Place it walks the ordinary path, so a gate can still refuse it, and the row will say so. Whatever you do, the agent follows the proposal on paper and records what it would have done, so the record grows even when no money moves.When it stops itself
This applies to an agent that uses a model. Three failed model answers in a row and the agent switches itself off: it tells you once, places nothing, and the bot’s other sources keep running. The strip and the Agent decisions tab show Stopped with the reason and a Resume button; saving the panel also resumes it.The daily report
- The agent’s day: once your day has ended, each bot with a decision gets one short paragraph — the numbers first, then what the skips had in common. It is a card at the top of the tab and a notification; no advice, no predictions.
The record
- Agent decisions tab on the bot: every look, skips included, who decided it (an engine rule, the screening model, or the deciding model), the reason, the levels, and Open the look to see exactly what it saw. Pip’s scan rows are engine rows and say “no agent credit used”.
- Signals tab: every look that spent credits, and every proposal once decided, as a row tagged agent look, proposal, you skipped, expired, you placed it, the agent placed it or refused, with the reason and the credits it cost. A signal the agent held for judging stays under its own source, tagged held for the agent.
- Agent skipped is its own status and filter. A decision not to trade is not a failed signal.
Your own agent
On Standard and up you can build agents of your own in Agents and run them on your bots. An agent of your own is a script on a bot’s chart, plus Pip, the assistant you talk to while you build it. Pip, the built-in agent, stays the default on every bot.- How many: Standard keeps 2 agents, Pro 4, Ultimate 10. Pip is built in and is not counted.
- On a bot: a bot can carry up to 5 agents. One position is open per bot at a time: when two agents see a trade, the first one in takes it, and the others skip and say why. The bot’s daily limits count every agent’s trades together.
- Signals only: like every source, your agent sends buy or sell. Position size, the Exit rule, Martingale, Filters, Schedule and Limits stay the bot’s.
Build it
Agents are made in the Build window: the chat with Pip on the left, the chart in the middle with the script under it, and the script’s settings on the right.1
Start a new agent
Open Agents and press New agent. It opens in the Build window. Pick a bot in the header: the chart shows that bot’s own candles, so you need a bot first.
2
Ask Pip, or paste a script
Say what you want on the chart in plain words and Pip writes the script, checks it on our engine and puts it in the editor. Or paste a TradingView-compatible Pine script of your own. The Script tab marks any line that will not run as written, and Fix with Pip rewrites just those lines.
3
Run it
Run draws the script’s signals and trades on the chart. Under It trades on, each kind of signal the script gives (an alert, a label, a shape, an order) is set to Buy, Sell or Ignore. Hover over a signal to read its full text.
4
Replay it
Replay in the header re-runs the script on history as the bot would have traded it: the bot’s size, ladder and Exit rule. Results are in money. Performance shows the totals, and Trades lists every trade with filters for Taken, Won, Lost and Passed over (signals that came while a trade was open). Runs keeps every Run and Replay with the script that made it.
5
Publish
Publish freezes the draft as a version. Bots run the version you publish, never a draft. Versions lists them, and you can go back to an earlier one.
6
Add it to a bot
In the bot builder, add it from the signal source picker. An agent must be replayed once before a bot can run it.
A replay is the script re-run on past candles. It cannot know the bot’s gates, the news filter, the schedule or real fills, so a live day can take fewer trades than the replay showed.
When the script fires
On the bot, the agent’s panel asks what happens to each signal:
At Suggest each signal arrives as a card for your tap; at Enter it is placed. Live, the script runs at every close of its timeframe on the bot’s own candles.
Tune
The Tune tab lets Pip replay the window and change the script’s settings as it goes, inside the bounds the script sets. A setting you pin is never moved. Your own fixed settings run on the same days beside it, so you can compare. Live tuning on a bot is a switch: when it is on, Pip tunes the bot every hour and after red news.What costs agent credits
Sharing
An agent can be shared to the Community or by link, and other traders can save a copy to change for themselves. A copy is a fork: your later versions never reach it unasked. Readers never see your broker, account, money or sources. Browse shared agents on the Agents page lists what others shared.Agents taught in words (classic)
Agents made before the Build window were taught with rules, graded looks and files instead of a script. They keep running as they are, and their rules are read only. On a bot they use a model on every look, so each look costs agent credits. Converting one to a script is optional: open it and choose Convert to script with Pip.Scripts without an agent
A script on its own lives under Scripts. It reads the candles and says buy or sell with no AI model and no agent credits, and it is free on every plan. Each trade it opens counts toward your plan.What it will not do
- Place a trade without your tap, at the Suggest tier.
- Trade a pair the bot does not have.
- Set a size, move a stop or a target, close a trade, or skip a gate. Those are your bot’s settings.
- Skip a sound setup because of the size your ladder would take next, the number of trades or losses today, or the cost of the trade. Those are your bot’s Position size, Martingale, Limits and Exit rule to set.
- Charge credits for a failed model call or a check the engine settled itself.
- Keep calling a model that keeps failing: three in a row and it stops and tells you.

