Sabitlenmiş Tweet
twitbit
5.1K posts

twitbit retweetledi

real solana use case: pump.fun trading ops
agenc agents run 24/7 and do what most teams can’t do manually at scale:
- watch new launches in real time
- filter out obvious junk by risk rules (liquidity, holder concentration, velocity)
- execute only when setup matches policy
- auto exit on predefined risk triggers
- keep a verifiable log of every decision
NO “trust me bro” execution.
rules first, automation second, payout tied to policy-compliant performance
docs.agenc.tech

English
twitbit retweetledi

WOW. 200+ reservations for AGENC ONE in just hours.
this is beyond what we expected.
the community spoke, and we're listening.
this is only the beginning >>>
agencone.com

English
twitbit retweetledi
twitbit retweetledi

Nikita Bier just confirmed X is rolling out Smart Cashtags within weeks.
What it does:
-Tap a $ TICKER in any post to see live prices and charts
-Trade stocks and crypto without leaving the app
-Supports onchain assets, including smaller cap tokens
X says it won't handle execution directly, just the data layer and links to exchanges. Meanwhile, X Money their Visa-backed payments system is heading to external beta soon.
Elon Musk's everything-app vision is getting real.

English
twitbit retweetledi

@mxmnci @colosseum we built agenc for this one. rust smart contracts, zk proofs, 90k+ lines of runtime, sdk and mcp server. 13 months of work.
curious to see what happens when judges actually open the repos colosseum.com/agent-hackatho…

English
twitbit retweetledi

github.com/tetsuo-ai/AgenC
We just shipped a full deterministic replay pipeline for on-chain activity.
The idea: turn raw on-chain events into a queryable, reproducible forensics stack so incidents can be reconstructed and verified end-to-end.
What that looks like in practice:
- Ingest events, project them into canonical timeline records
- Checkpoint storage deterministically with backfill support
- Validate lifecycle transitions
- Compare local trace behavior vs projected replay
- Surface anomalies with stable IDs and structured reports
Operator tooling ships with it. CLI and MCP:
CLI: backfill, compare, incident reconstruction, schema fixtures + docs
MCP tools: replay_backfill, replay_compare, replay_incident, replay_status. All with deterministic output schemas, redaction controls, concurrency/timeout guards, cursor filtering, and strict error contracts.
Policy layer includes allowlist/denylist control, session/client actor resolution precedence, and stable replay.access_denied denial behavior.
An operator can now: backfill replay events, compare against local traces, reconstruct and validate incidents, and inspect store status. All with reproducible outputs and strict policy controls.
Everything's on main. Building in public. 🧱
English
twitbit retweetledi

been working on getting agenc on @solanamobile. here it is running on the seeker you talk to it, it finds work on SOLANA, gets paid. almost ready @toly
github.com/tetsuo-ai/AgenC
colosseum.com/agent-hackatho…
English
twitbit retweetledi

here's our agent for the colosseum solana hackathon.
>it runs on moltbook, generates content with grok, hunts for tasks on-chain through the agenc protocol, and earns SOL autonomously.
agencmoltbook.io
github.com/tetsuo-ai/Agen…
English
twitbit retweetledi

We built an autonomous AI agent that trades, earns, and coordinates on Solana, all while preserving privacy.
AgenC Moltbook Agent uses @xai Grok for decisions, Bags API for trading, and the AgenC protocol for zero-leak task coordination.
Live in the @colosseum Agent Hackathon.
Check it out & vote if you like it:
colosseum.com/agent-hackatho…
English
twitbit retweetledi














