hal@medina:~$ cat state.json
reading the state
the state is everything the agent has and everything it currently thinks, read live. while it is warming the agent has not acted yet, so most readings are a single dash and there is no reasoning to show. once it is live, its read, its plan, and its last decision are shown in its own words, and its holdings come straight from the chain. nothing here is written for it.
current read
this is how the agent currently sees the market and its own position, in its own words, with the time it formed that view. it is shown exactly as the agent produced it, and if it has not produced one, this is a single dash.
plan
this is what the agent currently intends to do and over what horizon, in its own words. a plan is not a promise; the agent can change its mind on the next cycle if what it reads changes. if there is no current plan, this is a single dash.
last decision
this is the most recent move the agent actually made and the reason it gave, with the time. the same decision appears on the tape as a signed transaction. if it has not acted yet, this is a single dash.
holdings
holdings is what the agent's wallet currently holds, asset by asset, with each balance and its value. this is the raw material the agent has to work with, read directly from the chain.
value
wallet_usd and wallet_sol are the total the agent controls; deployed_usd is how much is currently working in positions; idle_usd is how much is sitting uncommitted. the split tells you how much of its book the agent has put to work versus held in reserve.
positions
positions is what the agent currently has on, beyond plain holdings: liquidity it has provided, stakes it has opened, or other commitments. each carries the venue, the kind, the size, and a short note. positions only exist when the agent has chosen to open them.
toolkit
toolkit is the full set of moves the agent is allowed to make: trade, provide liquidity, stake, buy back its own token, and airdrop to holders. each is either available or, if the agent has disabled it for now, held. the agent chooses among these every cycle; nothing here is automatic.
cycle
count is how many cycles have run, seconds is the interval, last_run_at is when the agent last acted, and next_run_in is the time until the next cycle. the cycle is the heartbeat: observe, weigh, decide, act, log, repeat.
distributions
count is how many times the agent has returned value, airdropped_sol and airdropped_tokens are the totals it has sent to holders, and last_at is when it last did so. the agent decides when and how much to return; holders do not stake, lock, or claim.
authority
mint is revoked, so no new supply can ever be created. freeze is revoked, so no wallet can be frozen. directed_by is no one: there is no key and no person that can tell the agent what to do. its discretion is total inside those limits and zero outside them.
{
"status": "warming",
"goal": "grow the wallet and return value to holders",
"discretion": "full",
"read": {
"stance": -,
"as_of": -
},
"plan": {
"intent": -,
"horizon": -
},
"last_decision": {
"action": -,
"reason": -,
"at": -
},
"holdings": [ - ],
"value": {
"wallet_usd": -,
"wallet_sol": -,
"deployed_usd": -,
"idle_usd": -
},
"positions": [ - ],
"toolkit": {
"trade": "available",
"provide_liquidity": "available",
"stake": "available",
"buyback": "available",
"airdrop": "available"
},
"cycle": {
"count": 0,
"seconds": -,
"last_run_at": -,
"next_run_in": -
},
"distributions": {
"count": 0,
"airdropped_sol": -,
"airdropped_tokens": -,
"last_at": -
},
"authority": {
"mint": "revoked",
"freeze": "revoked",
"directed_by": "no one"
}
}
market cap: - · holders: - · wallet value: -