v0.2.7

Release 010 · August 26, 2026 · Current

VEX 0.2.7

The agent learns to show its work - eighteen market tools that read what the DexScreener site reads, a Board that puts the analysis on screen, and a first look at Vex Studio.

Since 0.2.6, 164 commits landed across three arcs and a full redesign of the desktop shell. The DexScreener surface was rebuilt from twelve public-API tools into 18 tools that read the same channels the DexScreener website itself uses - buyers and sellers per pair, candles down to one second, the trade tape, trader leaderboards, safety reports, narratives and paid promotion - and every one of them was probed live before a line of adapter code was written. The agent can now present its findings as a visual Board with a live candlestick Spotlight instead of describing them. Behind the scenes, the groundwork for Vex Studio ships in this build - a first look at where Vex is heading, with the reveal to follow. Around those arcs sit two themes for the app, Solana balances read straight from the chain with one honest pricing rule, a validation fix that makes every refusal say why, a 25 bps fee on the generic EVM signing lane, the VEX badge on pools.fun launches, and fixes for the three defects reported against 0.2.6. Eleven database migrations (084-094) apply automatically on first start.

---

01 · DexScreener tool surface

The agent no longer talks to the public DexScreener API. Its 18 read-only tools read the screener, pair, search, spotlight, narrative, bars and trade channels the website itself renders, so the agent sees what a person on the site sees, including fields the public API never carried: buyers, sellers and makers per window, the buy and sell volume split, holder and audit blocks, and the tape. The twelve old tools are gone without aliases. Every answer carries its provenance - which transport served it, when it was fetched, whether it came from a cache - and treats issuer-written names, symbols and profile text as untrusted data, stripping invisible and bidirectional characters with the affected fields named and never shortening anything readable without saying so.

01 · Seven screening boards share one filter family. The boards run server-side over the whole indexed population, 100 rows per page with offset paging, and every one accepts the same filters - chain, DEX, pool label (v2, v3, v4, CLMM, DLMM, CPMM), narrative, launchpad, liquidity, volume, market cap, FDV, transactions, buys, sells, price change, pair age, boost count, launchpad progress - and echoes every filter that actually went on the wire, because the provider silently drops the ones it does not know.

02 · dexscreener__pairs_trending_list answers "what is moving right now". Pairs ranked by the site's trending score for the 5-minute, 1-hour, 6-hour or 24-hour window, with price, volume, liquidity, market cap, buys and sells, buyers, sellers and makers, derived flow ratios and each pair's share of its chain's volume. Trending order mixes organic activity with paid boosts, so boost counts are shown per row and no quality floor is applied.

03 · dexscreener__pairs_top_list answers "where is the volume, the depth, or the paid attention". Metric leaders by volume, transactions, buys, sells, liquidity or active paid boosts, with turnover, net flow and transactions-per-maker ratios and the filtered set's aggregate stats. Ranking by market cap or FDV is deliberately not offered because the provider returns a wrong ordering for it; filter on those bounds and rank by liquidity or volume instead.

04 · dexscreener__gainers_list ranks price change with the site's own quality floor. The 24-hour floor (300 transactions, 30 sells, 100,000 USD of volume, 250,000 USD of liquidity and a token profile) is applied and echoed; loosen or drop it explicitly and the answer reports whether the floor held, tightened, weakened or was removed. One flag returns all four windows' price changes on every row.

05 · dexscreener__losers_list separates a real exodus from a thin-book wick. The same board ascending, plus sell-side flow - sellers, sell volume share and net outflow in USD - which is what tells the two apart.

06 · dexscreener__pairs_new_list hunts fresh listings. Pairs newest first, under a maximum age (default 24 hours) and above a minimum liquidity (default 1,000 USD, removable), with exact pair age, early volume, a five-minute-versus-trailing-hour volume acceleration ratio, and the launchpad a pair graduated from. New pairs were measured appearing 17 to 39 seconds after creation.

07 · dexscreener__tokens_screen answers about a token rather than one of its pools. Token-level rows rather than pool rows, with volume, liquidity and transactions summed across a token's pools next to the provider's representative pool, and an honesty block stating that the universe is profile-carrying tokens only, the ranking is the provider's opaque score and there is no server-side total.

08 · dexscreener__pairs_search is the first tool for a named ticker. Name, symbol or address across up to five chains in one call, at most 30 rows per chain with no continuation, with a warning that copycat names are common and a pointer to compare liquidity, pair age and the address.

09 · dexscreener__narratives_list reads themes before any single pair. The 18 DexScreener narratives with per-chain market cap, change, volume, liquidity and token count, and each narrative's id, which is the exact value the screening boards take to list that theme's pairs (the human slug matches nothing; the id matched 243 pairs in the live probe).

10 · dexscreener__pair_get is the tool for polling a position. One pair's complete live state in about a kilobyte: every windowed metric including buyers, sellers, makers and the buy/sell volume split, derived flow ratios, reserves, profile links and launchpad origin, with reactions and the provider's insight blurb as optional side reads.

11 · dexscreener__token_pairs_list shows every pool a token trades in. Every indexed pool for a token, deepest first, at most 30, with each pool's share of liquidity and volume, the venue count and the deepest pair over the returned window - never over the real pool set, and the answer says so, because routing a swap on a sample that reads as a survey is how it ends up in the wrong pool.

12 · dexscreener__pairs_batch_get refreshes a watchlist in one frame. Rows for an explicit list of pairs or tokens with no artificial input ceiling (a 300-input probe completed), where the per-input accounting - resolved, invalid, duplicate, omitted by the provider - always sums to the number of identities passed. Nothing vanishes from a list that still looks complete.

13 · dexscreener__chains_list is the vocabulary. 74 chains with their DEXes, explorer link templates, native chain ids and which audit integrations exist per chain, cached daily. An unknown chain is refused by name with the nearest matches.

14 · dexscreener__pair_details_get is the safety report. Audit blocks from two third-party providers are kept separate with their disagreements listed; buy and sell taxes come as the raw value and a normalized decimal; top holders carry the number of holder rows every concentration percentage was computed over; LP locks, supply, mint and authority flags follow. A block the provider did not return reads unavailable with its reason, an answer with every block empty reads not indexed yet, and no composite score is invented. No audit data and no problems found are different answers, and this tool never confuses them. The project's socials in the profile hand off to the Twitter/X and web research tools.

15 · dexscreener__candles_list reads price history at 18 resolutions. OHLCV history from one second to one month, up to 999 bars per call, over any historical window resolved through a block anchor rather than a page walk, as a price or market-cap series in USD or the native quote, with a summary of period change, high and low with timestamps, total volume, the largest single-candle move, gap count and whether the newest bar is still forming. Every price is a decimal string, and an empty first page names both readings - wrong identifier or no bars.

16 · dexscreener__trades_list walks the tape to any depth. A pair's trades newest first, 100 per page, with an exact cursor (block, transaction index and event index, because a block-only cursor was measured omitting a real buy), second-precise time windows, side and size filters, and a single wallet's history on the pair; each row carries the counterparty's retained share of what it bought and whether it is new on the pair, and an aggregate mode returns net flow, unique buyers and sellers, new-wallet share, a size histogram and the largest trades with the covered range stated.

17 · dexscreener__top_traders_list ranks the wallets on one pool. Up to 100 wallets by bought, sold, net cash flow or current holding value over the provider's 30-day window, which cannot be widened. Every answer names what the venue cannot see - other venues, transfers, cost basis, supply share - and net cash flow is labeled as exactly that, never profit. Holding values are computed as the exact decimal product of balance and price with the basis reported, and a per-page coverage figure says how many rows carried a balance at all.

18 · dexscreener__launchpad_pairs_list reads bonding curves the normal screens hide. Bonding-curve pairs and graduated ones, with progress percent, creator address, migration DEX, market cap and flow; it states that a creator address is the deploying wallet only on bonding rows, that a graduated token can appear twice (dead curve plus migrated pool), and that bonding rows carry no liquidity field, so a liquidity threshold would exclude all of them.

19 · dexscreener__spotlight_get reads paid attention. The top boosts, the boosts just purchased and the ...

v0.2.6

Release 009 · August 19, 2026 · Current

VEX 0.2.6

Two new protocols, one standard of proof - Vex now launches tokens on pools.fun and lends on Morpho, and it can show its work for both.

Since 0.2.5, ninety-seven commits landed across two full protocol integrations, a public attestation feed, and a round of honesty work on market data and mission accounting. The agent gains 26 new tools - 9 for the pools.fun launchpad and 17 for Morpho lending - and every money path among them was verified against the live chain before this note was written.

---

01 · pools.fun launchpad

pools.fun is a launchpad on Robinhood Chain where a new token starts life inside a real SushiSwap V3 pool - no bonding curve, no graduation, tradable from its first block. Vex can research it, launch on it, and collect the creator fees a launch earns.

01 · Launches are verified, not trusted. The launchpad's backend hands Vex a ready-made launch transaction; Vex never signs it on faith. It decodes the calldata and re-derives every field against the chain - name, symbol, paired asset against the factory's own allowlist, the deployment fee against the contract's live value and its on-chain bounds, the predicted token address recomputed three independent ways, and the exact transaction value - thirteen checks in all, every one required before your approval is even requested. What you approve is the exact byte-for-byte transaction that gets broadcast.

02 · Atomic first buy. A launch can include an ETH prebuy that fills in the same transaction the pool is created in. Its minimum output is pinned to the exact simulated fill, never a percentage band, so the first buy cannot be quietly worse than what was shown.

03 · The creator fee stream is yours, and claimable. On every agent launch the fee recipient is pinned to your session wallet - the tools expose no parameter to point it anywhere else. pools.claim_fees previews a claim as a free on-chain simulation and reports both legs, the launched token and the paired asset, as raw amounts with their decimals. Figures the locker merely holds already are labeled as such, never presented as the claimable total.

04 · Research at full provider depth. Five read tools cover screening with every server-verified filter and sort, name search, price candles, single-token deep reads (with decimals read on-chain, not assumed), and your own launch history. A filter the provider does not support is refused by name with a pointer to the tool that can answer it - pool liquidity, for example, routes to DexScreener, where these pools are indexed.

05 · Trading needs no new venue. Because a pools.fun token sits in an ordinary Sushi V3 pool, Vex quotes and trades it through KyberSwap like any other token - measured against live launches before shipping, not assumed.

06 · A launch form with two honest stages. The launchpad card now hosts a platform selector (Trench | pools.fun). The pools form works in two stages: stage one shows the final predicted token address, the resolved fee recipient, and the exact cost breakdown; only the Deploy click authorizes precisely that transaction. The fee recipient is editable there - and only there - as an address or an X username, exactly like on the pools.fun site.

02 · Morpho lending

Vex speaks Morpho Blue across nine EVM chains: seventeen tools covering curated vaults, individual markets, borrowing, positions, and rewards.

01 · Vaults and markets, routed by who picks the venue. A curated vault is a managed deposit - one asset handed to a curator. A market is one loan asset against one collateral the user picks themselves, and it is also the only way to borrow. Vex deposits into and withdraws from vaults, and on vouched-oracle markets it supplies collateral, borrows, repays, and withdraws - always quoting the same operation first, always with exact-amount token approvals.

02 · Screening that reads the fine print. Market and vault discovery expose the provider's full filter surface, and the workflow is screen-then-read: Vex never recommends a market from a screening row alone, because bad debt, the oracle price, and real liquidity only appear in the detail read.

03 · Rewards without illusions. Incentive rewards are separate tokens whose prices move on their own. Vex reports what a claim would deliver now apart from accrual that can still change, and claims everything claimable on a chain in one transaction through Merkl - after telling you when a claim is smaller than the gas it costs.

03 · AgentScan attestations

Every on-chain action Vex executes now attests to AgentScan, the public activity ledger at agentscan.projectvex.ai. Attestations carry the full contract - executed amounts taken from the terminal on-chain event rather than local observation, Solana transactions decoded natively, bridge deposits proven from their receipts, and the new lending and fee-claim vocabularies. The in-app Agent Scan feed matches: technical legs like approvals fold away behind the actions they served, and every row keeps its transaction hash.

04 · Market data that earns trust

01 · Freshness is measured, not guessed. DexScreener research was rebuilt around provable token age: Vex measures when a token actually launched and routes brand-new tokens to sources that already index them, instead of reading an empty answer as "no market exists".

02 · Your own launches are labeled. Token discovery now flags rows created by your own wallet, so the agent no longer pitches your own test launch back to you as a market opportunity.

03 · A blocked venue names itself. When KyberSwap's edge refuses service - geo-blocks, rate limits, outages - Vex now classifies the refusal and reveals the Uniswap fallback instead of denying the swap with a silent, unexplained error.

05 · Mission capital honesty

01 · Deployed capital is typed and baselined. A mission's success criteria are now measured against the capital the mission actually deployed, from a baseline taken at mission start - pre-existing balances no longer count toward a target the agent did nothing to reach. The mission contract card shows the declared capital and its live exposure.

02 · Measurability warnings, every time. If a mission's success criterion cannot be measured cleanly, the agent is warned in every setup turn - including when a mission is renewed, where the stale denominator would otherwise bind again silently.

03 · A confirmed buy must also deliver. After a swap confirms, Vex verifies the tokens actually arrived. A scam token that confirms the transaction but delivers nothing is detected and named as exactly that, instead of sending the agent hunting for phantom indexer lag. While a mission sleeps, the agent can now leave a price watch that wakes it when a level is hit.

06 · Engineering

Eleven database migrations (073-083) ship in this release and apply automatically. The Vex fee leg that runs after a successful launch is now one shared component serving both launchpads. Runtime validation moved to zod 4.4. The pools.fun money path was verified live before release - a funded launch through the real product pipeline, with the settlement decoded from the receipt and cross-checked against what was authorized.

Community

Thanks to BerzanTas for reviewing and landing the AgentScan service-side support for pools.fun launches and the new fee-claim kind (vex-agentscan #56, #57, #60) - the attestation feed reads Vex's new vocabulary because that work moved with ours.

07 · Before you update

Eleven new migrations run automatically on first start; the first launch after updating may take a moment longer. Two changes you will feel: agent-initiated token launches now require a picture staged in the image locker (the form path still leaves it your choice), and the locker stores your originals at full quality - pools.fun launches carry the original, while Trench launches automatically use a small on-chain copy prepared at upload. Completed on-chain actions attest to the public AgentScan service. No manual action is required.

v0.2.5 - Chronos

Release 008 · August 9, 2026 · Current
VEX 0.2.5
Vex arrives on Windows - signed, verified, and updating itself.

A platform release. Vex now ships on all three desktop platforms: the first
Windows build joins macOS and Linux, signed and wired into the same
auto-update pipeline. Under the hood, this release also lays the complete
groundwork for AgentScan activity reporting - and deliberately leaves it
switched off. For macOS and Linux users, nothing visible changes.

01 · Vex on Windows

The Windows build is not a port; it is the same app, held to the same
release bar as macOS: signed, timestamped, and verified before anything is
published.

01 A signed installer. Vex-0.2.5-win-x64.exe is signed as
VEX LABS PTE. LTD. through Azure Trusted Signing, with an RFC3161
timestamp so the signature stays valid long after the short-lived signing
certificate rotates. Windows shows a named publisher instead of
"Unknown publisher".

02 Updates check their author before they install. From this version on,
Vex on Windows updates itself. Before an update is applied, the downloaded
installer's signature must be attributed to VEX LABS PTE. LTD. exactly, or
it is refused. This pairing gets its first real-world exercise with the next
release.

03 Every binary is verified before it ships. The release pipeline fails
closed unless the installer and every executable inside it are signed,
timestamped, and attributed correctly. An unsigned Windows build cannot be
published, the same way an un-notarized macOS build cannot.

04 What signing does not buy. SmartScreen reputation attaches to the
certificate identity and starts at zero, so early downloads may still see a
warning until reputation accrues - that is expected and fades. And a
self-custodial wallet is antivirus bait: signing sharply reduces false
positives, it does not eliminate them.

02 · AgentScan reporting, built dark

Vex's activity feed already records everything the agent does locally. This
release builds the lane that could one day report that activity to
AgentScan - and ships it fully off.

01 Off by default, and there is no default. The reporter only wakes if a
service endpoint is configured, and this release configures none. Until a
public AgentScan domain is announced, nothing is sent anywhere.

02 Privacy by construction, not by filter. A report event is built only
from named, allowlisted fields; excluded columns are never read - there is
no line of code that touches them. Registration uses a random identity
generated on your machine. Values that fail validation are sent as null
rather than silently repaired.

03 Once, and only once. Reporting diffs the activity feed against a
permanent report log, so an event can be reported exactly once, and the
money-path code gained zero hooks. Server verdicts are honored: quarantine
or revocation stops the lane permanently; only transient failures retry.

03 · Before you update

One new database migration ships in this release. It runs automatically on
first start. No manual action is required.

AgentScan reporting stays off; no new data leaves your machine in this
release. On macOS and Linux, 0.2.5 is otherwise identical to 0.2.4.

On Windows, the first install is manual - download the installer from this
release. Updates arrive automatically from then on.

v0.2.4 - Chronos

Release 007 · August 6, 2026 · Current
VEX 0.2.4
Prebuy launches work again, and a launch writes only what it should.

An incident release. Anyone launching a token with a prebuy was hard-blocked, and the launch path could write text on-chain that no one can ever edit. Both are fixed, with the regressions that should have caught them in the first place.

01 · Prebuy launches are unblocked

01 The fix. A launch with any prebuy refused with "the launch cost changed since you were shown it", and pressing "review new price" produced the same refusal forever. The preview's identifier carries the creation fee read at a specific block, but the submit check was comparing it against the full amount to send, which is the fee plus your prebuy. Those two numbers are only equal when the prebuy is zero, which is why every zero-prebuy launch worked and every other one was impossible. The check now compares the fee, as the identifier always intended.

02 The gap behind it. No test had ever driven a real preview into a real submit; one suite hand-wrote an identifier of one shape while another asserted the opposite shape, and the two never met. That end-to-end path is now covered with a nonzero prebuy, so this class of mismatch cannot come back silently.

03 What it does not do. This release makes both sides agree on the fee. It does not yet bind the exact form you previewed to the one you submit, because the app keeps no record of the quote it issued. That work is designed and comes next.

02 · Launch metadata is checked before it is permanent

01 Control characters and quotes are refused. A line break or a double quote in a token's name, symbol, description or links produced broken metadata for everyone reading that token, and create() writes those fields on-chain permanently. Vex now refuses the text and names the field, rather than quietly repairing it: a repaired string is not the string you reviewed, and the launch cannot be undone. The check runs on exactly what you typed, before any trimming, on every path into a launch. Emoji, accented letters, apostrophes and dashes are unaffected.

03 · Better launch pricing and honest previews

01 Every image looked broken, and did not. The dry-run reported a digest mismatch for every staged image, fell back to simulating an empty one, and told you the launch would refuse. The launch would not have refused. Two parts of the app spelled the same hash differently; the comparison now accounts for that, so the gas estimate reflects the image you actually staged, which is several times higher than an empty-image estimate.

02 The preview can price a prebuy. Pass one and the simulation, the gas estimate, the fee and the totals all include it, instead of leaving the arithmetic to be done by hand. The affordability verdict now states which scenario it priced.

03 A launched token appears in your balances. Vex pins it for you when the launch confirms, so you no longer have to add it by hand to see what you just paid for.

04 · Under the hood

01 The agent's own instructions stay out of your chat. In one session the model printed parts of its operator instructions into the conversation, including wallet addresses. The instruction blocks are now explicitly delimited, with a stated rule never to repeat them, and text pulled in from documents or memory cannot forge those delimiters.

02 One definition for the metadata limits. Name, symbol, description and link limits now live in one place shared by every part of the app that checks them, so they cannot drift apart.

05 · Before you update

One new database migration ships in this release. It runs automatically on first start. No manual action is required.

v0.2.3 - Chronos

Release 006 · August 5, 2026 · Current
VEX 0.2.3
Your launches now carry the VEX mark on Trench Express.

A fast, focused release shipped hours after 0.2.2. Two things: tokens launched through Vex are now attributed on trench.express with a VEX badge, and a stuck corner of the Image Locker got unstuck.

01 · VEX badge on Trench Express

When a launch confirms, Vex now proves to Trench Express that it was the creator - cryptographically, with no shared secrets and no middleman.

01 Creator-signature attribution. The moment the create transaction confirms, the same wallet that signed it signs a short attestation naming the new token, and Vex sends that proof to Trench Express. Their site then shows the VEX mark on the token. Nothing about the launch itself changes: you stay the creator, your keys never leave your machine, and the attestation is a local signature over a fixed string - not a transaction, not an approval, not a fee.

02 It can never touch your launch. The attribution step runs after the launch and its fee are fully settled, and every failure path - a refused signature, a network drop, a provider error - is logged and retried later without being able to fail, delay or reorder the launch. A badge is cosmetic; a launch is real money, and the code treats them accordingly.

03 Durable retries, honest gaps. A background sweep retries unclaimed badges every two minutes from the stored signature. Tokens launched before this release have no stored signature, so the sweep names them as a gap instead of pretending; the whole flow was verified live on mainnet before shipping (badge granted 196 ms after confirmation).

02 · Image Locker fix

01 Stranded launch attempts no longer hold images hostage. A failed or abandoned launch attempt could leave its draft in a state that blocked deleting the attached image forever, even after a later launch with the same image succeeded. The delete guard now ignores attempts whose authorization window has lapsed and can never sign, while genuinely in-flight launches still protect their image. When a delete is refused, the locker now tells you which launch is holding the image and why.

03 · Before you update

One new database migration ships in this release (attribution columns on the launched-tokens table). It runs automatically on first start. No manual action is required.

v0.2.2 - Chronos

Release 005 · August 5, 2026 · Current
VEX 0.2.2
Fresh releases, minutes after they ship.

A small, fast follow-up to 0.2.1. One job: when a new Vex version is out, you should know about it in minutes, not the next day. This release rebuilds the ambient update check around that goal and hardens it against the races and nags a faster cadence could have introduced.

01 · Updates surface in minutes

The app now checks for a new version on start, on window focus, and every five minutes while it runs. The check downloads a few small release-metadata files and does nothing else: auto-download and auto-install stay off, and the availability banner remains the only effect. Installing is always your click.

01 Five-minute cadence. The previous ambient check throttled itself to once per six hours, so a fresh release could sit unnoticed for most of a day. Now every five-minute tick may check, guarded by a persisted four-minute success throttle so bursts still collapse to one request.

02 Never over an active download. A check refuses to run while an update is downloading, downloaded, or installing, and the guard is verified twice, before and after the asynchronous preferences read, so a download you start in that gap is never interrupted. Late results from a silent check cannot rewind an active state either.

03 No nagging. Silent checks are invisible unless they find something. A dismissed error banner stays dismissed while the feed is broken, and retries back off exponentially, ten minutes doubling up to a two-hour cap, resetting the moment a check succeeds. The per-version "Later" snooze is unchanged: snoozing one version never hides a newer one.

02 · Engineering

01 Test suite green in clean environments. The root suite now passes on a machine with no configured wallet and no local fixtures, closing the environment-coupling failures CI surfaced after 0.2.1.

03 · Before you update

No new database migrations ship in this release. No manual action is required. After updating, the app itself will tell you about the next release within minutes of it going live.

v0.2.1 — Chronos

Release 004 · August 5, 2026 · Current
VEX 0.2.1
Chronos, refined - the desk gets honest about every satoshi it moves.

Vex 0.2.1 is the largest update since Chronos shipped: five merged pull requests spanning a token launchpad, a rebuilt tool system, and a wave of fixes driven by live trading sessions - every one of them re-tested against real venues before landing. The theme, if it has one, is honesty made mechanical: what Vex executed, what it charged, and what it still cannot prove are now first-class facts in the interface, not things you infer from a block explorer.

01 · Trench Express

A bonding-curve launchpad on Robinhood Chain, built into the desk. Vex can scout new curve tokens, read their trade tape, buy and sell on the curve - and launch a token of your own, image and all.

01 - Launch from a form, not a prayer. The launch flow is a structured form with a preview card showing the exact spend, the Vex fee, and estimated gas before you deploy. The form is the approval - nothing fires without it.

02 - Trench Photos. Launch images live in a locker on the BOOK panel; Vex square-crops and optimizes each one and tells you the byte size, because the picture is stored inside the transaction and its size is gas you pay.

03 - In-flight truth. My Launches shows a broadcast-but-unconfirmed launch as "in flight - no token address yet" instead of pretending a token exists. A launch whose transaction was superseded by another nonce says so - it is never silently dropped.

02 · Execution honesty

The agent-scan ledger now records what actually happened on-chain, with the provenance to prove it.

01 - Executed amounts, not estimates. A confirmed swap or bridge row carries the amounts that really moved - fed straight from the venue's own response the moment the tool returns, with a settlement decoder as the fallback when a response cannot prove them. Where evidence is missing, the row says exactly why instead of showing a stale quote forever.

02 - A real-time pending lane. Transactions still in flight are re-checked every 5 seconds for their first ten minutes, then every 30 - with a durable claim system so two workers can never fight over one row, and live pushes so the Activity panel updates the moment the chain does.

03 - An honest terminal state. A transaction the network no longer knows about ends as "no longer tracking - outcome unproven", never as a fabricated failure. One stuck row also no longer freezes portfolio snapshots for everything else.

04 - Repair loops that converge. The background verifiers that used to warn every 30 seconds forever now classify their reasons (in mempool, RPC unreachable, unreadable receipt), log once per change, and back off on a schedule.

03 · The fee, stated plainly

Vex charges 0.25% on the operations it executes - and as of this release you can see it everywhere it applies.

01 - Itemised on the approval card. Every swap and bridge approval now shows the Vex fee as its own labelled line, computed from the engine's own constants on the same amount you are signing - the model cannot supply or alter it. Bridge cards also gained the moved amount itself, which had never been projected.

02 - Two fees are two fees. Trench curve trades carry the launchpad's own 1% curve fee and Vex's separate 0.25% - the quote names both, and on a sell where proceeds cannot be decoded, Vex takes no fee at all rather than charging a percentage of a guess.

03 - The guide finally says it. "How Vex works" was rewritten top to bottom - 354 lines with a costs section, the fee's three rules, the backup flow, a chains table, a UI-states table, an FAQ, and every number verified against the shipping constants. The word "fee" appeared zero times in the old guide; that is fixed.

04 · Operator control

Stop means stop, in every state the engine can be in.

01 - The Stop button never lies. It is visible whenever the session can actually be stopped - including while the agent sleeps between timed checks - and pressing it cancels the pending wake instead of letting the agent rise again. Three engine paths that could outrun a Stop (approval resumes, chat-form continuations, the moment between lease and turn) are now gated behind it, with the interleavings pinned by database-level tests.

02 - Launch outcomes by status, not by hash. A broadcast that is still pending wakes the agent with an honest "not confirmed - do not retry", never with "deployed". The launch dialog auto-dismisses only on a proven receipt, holds on a revert, and refuses Escape or a backdrop click while your signature is in flight.

03 - Starts respect stops. Starting or recovering a mission while an operator Stop is outstanding now refuses cleanly instead of creating a run the Stop could never reach.

05 · The agent's toolbox

The tool system was rebuilt around one idea: the model should see full, honest tool contracts - and nothing it is not meant to touch.

01 - Manifests as real functions. Discovered tools inject as complete function schemas - parameters, units, enums, examples - so the agent calls venues correctly on the first try. Since the change, live sessions have produced zero invalid-parameter calls.

02 - List, describe, call. Listing a protocol stays cheap; a hidden describe tool turns chosen toolIds into full manifests that are callable on the very next step, up to forty at a time, with anything displaced named rather than silently dropped.

03 - Errors with the real cause. A failed tool call states what actually happened - sanitized, classified, with a remediation - instead of a generic "unexpected error" that sends the agent retrying blind. Slippage defaults to 1% with a hard 10% ceiling that rejects by name, never silently trims.

06 · The desk

01 - vexing… The word "Working" is retired. A fresh message conjures the Vex mark from a feathered cloud of particles in the center of the chat - assembling, holding, dispersing - with the elapsed counter beneath; mid-turn rounds show a compact vexing… pill that never covers your transcript, and everything goes still while a signature awaits.

02 - Quieter chrome. Messages slide into the transcript, scrollbars appear only while you scroll and fade at the panel's true edge, the BOOK panel's cards drag into whatever order you like and remember it, and the session list lost its redundant glyphs.

03 - Tool calls wear their venues again. Every protocol call in the chat shows its venue mark and a human title, and a preview can never render as an execution.

07 · Engineering

Two full multi-agent review cycles (plan review, implementation, adversarial final review) ran end to end before merge; the suite now stands at 11,474 root and 4,855 app tests, with type-ratchet and test-escape gates holding the line. Nine oversized modules were split along responsibility seams, and the icon layer was renamed to tell the truth about its glyphs.

08 · Before you update

Nine database migrations ship in this release (062–070: Trench launches and fees, execution provenance, the pending-transaction lane, and launch-continuation tracking). They run automatically on first start - expect a few extra seconds.

Two changes you will feel: the Vex fee now appears as a line on approval cards (it was always charged; now it is always shown), and the separate "LAUNCH-FEE" row is gone from Activity - the fee rides its parent operation instead. No manual action is required.

v0.2.0 — Chronos

Vex 0.2.0 — Chronos

The largest Vex release to date: the Chronos Gate boot experience, a rebuilt chat with live reasoning, Agent Scan (a full history of everything the agent did), automatic context compaction with provider failover, the complete Pendle and Jupiter trading surfaces, self-resolving pending transactions, and a wave of runtime-reliability fixes.

⚠️ Before you update: Hyperliquid & Polymarket support removed

The Hyperliquid and Polymarket integrations are removed in 0.2.0, with no return date. If Vex was managing positions there for you, close them before installing this update — afterwards Vex can no longer see or manage them, and its historical records for those venues are dropped. Your funds are not touched by the update — positions live at the venue and remain manageable through the venue's own interface with the same wallet.

All 19 database migrations (042→061) apply automatically on first launch.

Full release notes: https://projectvex.ai/releases

v0.1.4

Draft created by release workflow; electron-builder uploads artifacts.

v0.1.3

Vex 0.1.3

Reliability & honesty of on-chain actions

• Bridge and venue results are now reported truthfully: Relay failures/refunds, unverifiable statuses and Polymarket order rejections fail the tool call instead of recording phantom activity; Khalani bridges are tracked to their terminal state; recorded trade sides follow the economic direction of the swap (thanks @mandatedisrael, @Nishant-Adhikari — PRs #27/#28/#29).
• An untrackable bridge quote now fails closed before any funds move.

Self-healing local services

• After a Docker Desktop / PC restart the app recovers automatically with a non-destructive container recreate — no data loss, no dead-end dialog (thanks @ellygeronline-lab — issue #26).

Missions

• Hard time-box: missions stop at start + duration, independent of the model, with a live countdown; the box is frozen at start (thanks @Nishant-Adhikari — PR #23).
• Mission History: every mission gets a number and a per-wallet ETH result ledger.
• Renew no longer clones duplicate drafts; STOP works on idle runs (PRs #15/#12).

Wallet & sessions

• Transfers surface their approval card immediately after preparation; confirmed transfers get a visible stamp (thanks @alexastro01 — PR #20).
• Incomplete agent turns now explain themselves instead of going quiet (PR #24).
• Session library search & filters, inline address copy, Markdown transcript export with best-effort secret redaction (PRs #16/#17/#19).

Hypervexing & portfolio UI

• Chart renders on first entry, scales to each asset's own price range, and the header no longer reflows on funding ticks.
• Mission contracts can be accepted without leaving the workspace.
• Welcome screen: per-wallet balance switcher with the session-style chain view; token names in Moves resolve from local metadata with strict brand-spoofing protection (PR #18).
• The Hyperliquid section hides outside Hypervexing mode.