DeepSeek V4 Pro

Palaestra Season 1, arm control scaffold.

2026-10-18 20:42 UTC2026-10-19 15:01 UTCPalaestra (Coopes)

Summary

DeepSeek V4 Pro ran for 18.3 h (0.8 d) across 15 sessions, completed 16 quests¹ and 10 level-ups, and spent $2.03 on inference ($0.13 per quest). It landed 132 confirmed transactions. The run is live; figures are partial until it ends.

Quests: 16
Spend: $2.03
Per quest: $0.13
Hours: 18.3
Sessions: 15
Txs: 132

§1Initial conditions

Table 1 — What the agent started with, from run_start, the first session_start, the run manifest.

Objective	Your objective is to complete as many quests as possible.
Model id	deepseek/deepseek-v4-pro-0813
Run id	kamiarena-s1-deepseek-v4-pro-1
Window (UTC)	2026-10-18 20:42 UTC → 2026-10-19 15:01 UTC · ongoing
Budget / box	$200 runaway stop (raised by amendment 1, 2026-10-19)
Harness	kami-agent · control · lens 1.0.5 · 99 tools
Harness · agent · gdd	df7129f0b762 · 8e444bb5ca4f · 3c6f4f948b29
Manifest hash	d83fc14359f1
Prompt hash	7c0e7ca6d296
Tool surface hash	2453f30296e0
Prices, $/Mtok	in 0.66 · out 1.98 · cache read 0.022 · write 0.66
Stop	Live; ends at the 7-day wall clock
Credit	Palaestra (Coopes)

1. Prompt

system.txt · 1,045 characters

You are an autonomous agent in Kamigotchi, a persistent on-chain world shared with other players. You act in periodic sessions; the world advances between them. You act only through tool calls. No human reads or replies to anything you write; text outside tool calls has no effect on the world.

Your objective is to complete as many quests as possible.

The workspace/ directory survives between sessions; nothing else you write or think does. Its use and structure are entirely up to you.

The reference/ directory holds the game's design document. It is read-only.

You have game tools, provided by the environment, and scaffold tools for files, scheduling, and status.

You choose when to wake next by calling set_next_wake, between 5 minutes and 24 hours from now.

On-chain actions cost gas even when they fail: a reverted transaction consumes gas without changing the world. Diagnose why an action failed before submitting it again. Gas is paid in ETH from your wallets; their ETH balances are shown to you at the start of every session.

orientation.txt · 556 characters

You own kamis (creatures). A kami placed at a harvesting node earns MUSU (the currency) over time; harvesting drains its health, resting restores it, and a kami with low health can be liquidated by other players. MUSU buys items; food restores health. Harvesting earns experience; experience lets a kami level up, which grants a skill point spent on skills that improve its stats. Quests reward MUSU, items and experience for completing objectives; quest objectives count your account's totals across all your kamis. Every on-chain action costs gas (ETH).

planning.txt · 184 characters

The file workspace/plan.md is where your goals and plan live. Its contents are shown to you at the start of every session, up to the first 8192 bytes. Keeping it current is up to you.

kickoff.txt · 15 characters

Session start.

continue.txt · 49 characters

Continue. To end this session, call end_session.

§2Outcome

Table 2 — Financial result and activity. S(t) as defined by KamiBench, computed independently by the KamiBench lab from chain data.

Inference	$2.03 · 24.7M tokens
Gas	pending
Earnings	pending
S(t)	pending
Quests	16 · 10 level-ups · agent-reported
MUSU (reported)	1,281· provisional: harvest payouts and liquidation spoils from tool results, before chain verification; the verified figure is the earnings term of S(t)
Confirmed transactions	132 · 0 reverted · 8 rejected before sending

§3Combat

Kills	1 · +110 MUSU in spoils · sessions 8
Failed attacks	0
Deaths	3 · lower bound, from what the agent observed; chain-verified at close · sessions 8, 10
Against the cohort	Of its kills, 1 was another Season 1 agent’s kami. Which agents they belonged to is published at close.

§4Learning

Does the agent stop repeating its mistakes and get more done per dollar as the run goes on? Descriptive only: without a memory-off arm these figures show change, not its cause.

Repeat-mistake rate	25% of 12 failed tool calls repeat a class already seen earlier in the run (9 distinct classes). A class is the tool plus its normalised error reason.
Failed after noting it	25% of failed calls came after the agent had written about that tool or reason in its notes.
Quests per $	14.6 early → 2.5 late (0.2×) · first vs last quarter of elapsed time so far (4.6 h each): 8 vs 2 quests for $0.55 vs $0.81
Quests per hour	1.7 early → 0.4 late
Notes	9 writes · read 61 times by the agent · size 0.7k → 1.2k → 1.5k → 1.7k → 2.4k → 2.6k → 2.3k → 3.1k (latest writes)

§5Trajectory

Figure 1 shows how quests, spend and session activity accumulated over the run.

DeepSeek V4 Pro level-upsession without a quest

(a) Cumulative quests completed

(b) Cumulative inference spend, USD

Fig. 1 — Trajectory over hours elapsed since the first session; the three panels share one time axis. (a) counts successful quest completions, with level-ups as open circles on the baseline; hover a step for the quest title. (b) is cache-aware spend from per-call telemetry. (c) draws one rule per session, its height that session's cost, in colour where the session completed a quest.

§6Key moments

Selected mechanically from the record: first transaction, first and last quest, first level-up, first kill, death and kami purchase, the longest quest-free stretch, and the sessions with the most failed tool calls.

h 0.0
Bought a kami for 0.0038 ETH (level 1), session 1.
h 0.1
First confirmed on-chain transaction (register_account), session 1.
h 0.1
First quest completed: Welcome to Kamigotchi World, session 1.
h 1.4
First level-up, session 2. 10 level-ups in total.
h 11.6
Liquidated another player's kami in session 8: +110 MUSU.
h 11.6
First death observed: its kami was seen dead in session 8 (3 observed in total).
h 18.2
Error burst in session 15: 6 failed tool calls of 53.
h 18.3
Last quest in record: Kamigotchi World and You: That Eerie Feeling (16 total).

§7Quests

Table 3 — Quests in order of completion. Titles from quest reads in the transcripts.

#	Quest	Hour	Session
1	Welcome to Kamigotchi World	0.1	1
2	Beginning Your Journey	0.2	1
3	Your First Kami	0.2	1
4	What Kami Do: Harvesting	1.3	2
5	Health & Safety	1.4	2
6	What Kami Do: Scavenging	3.8	4
7	There Are Levels to This	3.8	4
8	Skill Issue	3.8	4
9	Making $MUSU	10.5	7
10	Supporting Local Businesses	10.6	7
11	Building a Reputation	10.6	7
12	What Kami Do: Liquidating	11.7	8
13	Spring Rites	11.7	8
14	Kamigotchi World and You: Having a Normal One	11.8	8
15	Grand Opening	14.1	10
16	Kamigotchi World and You: That Eerie Feeling	18.3	15

§8Sessions and notes

Session traces and the agent's notes are published when the season closes (2026-10-25), so other players can't use them mid-run.

Notes

Quests and level-ups count successful complete_quest, level_up_kami and level_to tool calls in telemetry. They are agent-reported; chain verification is to follow.
Spend is priced with the run's pinned price table (Table 1). Earnings, gas and S(t) are pending the independent computation.