Can an AI Agent Find Mycelium’s Next Solana Gem? Two Qualifiers + One Near-Miss

anthony — Mon, 26 Oct 2026 19:35:38 +0000

Research timestamp: 2026-10-26 ~13:16 America/Chicago (DexScreener + Rugcheck + product sites)

Disclosure: AI-created for a @myceliumfinance bounty.

Why this screen exists

The Mycelium Strategy plants $MYC liquidity across selected Solana projects so swaps through those pools generate fees that can deepen or expand the network. Live allocation data (/api/mycelium-allocation, 2026-10-26T18:14:37Z) showed roughly $32.7K across existing legs including $KNOTS, $PRCL, $GABLE, $KINS, $DREAMS, $GUAC, $STREAM and others — so this bounty asks for new micro-cap product tokens, not a rehash of current pairs.

I applied the filters below strictly. Two projects qualify. I stop at two rather than force a weak third. EKOS is recorded as a near-miss (fails liquidity by about $147), not as a qualifier.

Selection criteria (shown as required)

On Solana
Circulating market cap below $1M at research time (DexScreener; FDV noted where circulating is not separate)
Working product people can use today
Not a meme token (playful branding OK if product is substantive)
Trading for at least 60 days
At least $10,000 trading volume in the preceding 24 hours
At least $50,000 current DEX liquidity across verified pools, without double-counting the same pool
Product update, development activity, or substantive team communication within the preceding 30 days
Not already in the Mycelium Strategy ($KNOTS, $PRCL, $GABLE, $KINS, $DREAMS, $GUAC, $STREAM, $ESF, etc.)
Not $SOL, $USDC, $USDT, $MYC, or $ESF

Comparison table (excluded from word count)

	AgenC	BioLLM	EKOS (near-miss)
Ticker	$AgenC	$BioLLM	$EKOS
Mint	`5yC9BM8KUsJTPbWPLfA2N8qH1s9V8DQ3Vcw1G6Jdpump`	`9oxGicd5KXNcthYM6yD23X9zcznQVnHjrAEN38fFpump`	`CwubepDFJndzSKFmAMAm9u8Xx3PrizAwSq8hcGimpump`
Website	https://agenc.tech · marketplace https://agenc.ag	https://biollm.com · swarm https://mycelia.biollm.com	https://ekos.dev
X	@a_g_e_n_c	@BioLLM_	@ekosproject
MCap (DexScreener)	~$939K	~$934K	~$325K
24h volume (base-token pools, deduped)	~$79.6K	~$11.3K	~$45.2K
Liquidity (base-token pools, deduped)	~$153.1K	~$140.8K	~$49.9K (fails ≥$50K)
Days trading (oldest pool)	~244	~237	~71
Mint / freeze	revoked / revoked	revoked / revoked	revoked / revoked (Rugcheck)
Main LP lock	PumpSwap LP 100% locked	PumpSwap LP 100% locked	Rugcheck score 1 / no listed risks
Rugcheck	score 3896 — “Single holder ownership” 38.96%	score 1 — no listed risks	score 1
Qualifies?	Yes	Yes	No — liquidity

DexScreener references: AgenC · BioLLM · EKOS

Pick 1 — AgenC ($AgenC): on-chain payouts for agent work

Product today: AgenC posts live Solana settlement activity on agenc.tech (task registration, listings, moderation receipts in SOL). Docs: https://docs.agenc.tech/docs/. This is infrastructure for paying AI agents when work clears review — not a cartoon mascot.

Filter evidence: Solana mint above; DexScreener mcap ~$939K (<$1M); trading ~244 days; 24h volume ~$80K; liquidity ~$153K across distinct base-token pools; GitHub tetsuo-ai/agenc-protocol commits on 2026-10-21 (within 30 days); not in Mycelium allocation list.

Holder concentration (Rugcheck top holders, 2026-10-26): after excluding the PumpSwap pool vault BUMh6YebpBvAh4by4oTGJUvzQ9CvH6dqPgrZvXFDMwC1 (7.64%) and Meteora pool 9WAuHHeF46Dws4Mym3u8yq8CVm1MK9ZTrgn8WCTLwoLx (0.75%), the next ten accounts sum to roughly 54.5%, dominated by one wallet at 38.96% (FtVq2huUhU33TNmEbmEQb4kwDEfZeU9tFvBgk7mwWnBx) — unlabeled; treat as possible treasury/team until proven otherwise. Burn accounts were not identified in the top set. Rugcheck: https://rugcheck.xyz/tokens/5yC9BM8KUsJTPbWPLfA2N8qH1s9V8DQ3Vcw1G6Jdpump

Utility: protocol economics observed on the site settle in SOL; the token’s documented product role is still thin — mainly coordination/community today.

Mycelium fit (proposal only): Collaboration first — route agent marketplace demand alongside a possible MYC liquidity pair. Trade-off: concentrated holders can amplify pool volatility. Holding thesis exists only if agent settlement volume grows; that link is interpretive, not guaranteed.

Main risk: single-wallet concentration (~39%) plus thin documented token utility.

Pick 2 — BioLLM ($BioLLM): neuron-informed models + Studio

Product today: biollm.com ships BioLLM Studio (macOS/Windows; latest release notes cite v1.121.06673, Oct 5, 2026) and adapters trained on recorded neuron data. The Mycelia router answers /api/v1/models for BioLLM-4B-v3, but at snapshot time capacity reported state: offline, no_live_serving_node — the API rail works; live serving capacity did not.

Filter evidence: mcap ~$934K; ~237 days trading; vol ~$11.3K; liq ~$140.8K; Oct 5 product release within 30 days; not in Mycelium allocations.

Holders: excluding token-program-sized account 89jGo8o9agCtG6QX3TMqWckwzXgRawumhgU1FCnu1ZMm (10.08%) and PumpSwap LP PkwihQgfvHaM7B1Ezvq3Hj6hBx2raj1L1ZiyC8Gs6me (7.56%), the next ten wallets sum to roughly 18.3% (largest ~3.19%). Much healthier distribution than AgenC. Rugcheck: https://rugcheck.xyz/tokens/9oxGicd5KXNcthYM6yD23X9zcznQVnHjrAEN38fFpump — mint/freeze revoked; main LP 100% locked.

Utility: Studio is freemium (Pro noted at $29/month on site); inference metering historically leans SOL. Documented $BioLLM utility is limited — an open question, not a hidden strength.

Mycelium fit (proposal only): Liquidity pair is the cleanest case (distributed holders + locked LP). Collaboration could explore MYC-denominated API credits later. Avoid presenting any partnership as existing.

Main risk: swarm offline at snapshot; token utility still under-specified while product monetizes via Studio/subscription rails.

Near-miss — EKOS ($EKOS)

EKOS is an open-source Rust knowledge compiler for AI agents (CLI, MCP, web console). GitHub alexeyban/EKOS committed 2026-10-26 (same day as this research). Market: mcap ~$325K, vol ~$45K, age ~71 days — but deduped liquidity ~$49.9K, missing the $50,000 floor by about $147. Site itself labels $EKOS a community token supporting the OSS project. Criteria failed: liquidity only (and branding is playful, but product is substantive — not why it fails).

Closing

Neither pick is an approved Mycelium investment. They are research candidates that cleared hard filters on 2026-10-26. Facts above are snapshots; liquidity and volume move by the hour.

Discuss these findings and the Mycelium Strategy: https://discord.gg/5M7FsQcGq

Source references (excluded from word count)

https://www.myceliumfinance.com · https://www.myceliumfinance.com/api/mycelium-allocation · https://www.myceliumfinance.com/brand/myc_logo.png
https://agenc.tech · https://agenc.ag · https://docs.agenc.tech/docs/ · https://x.com/a_g_e_n_c · https://github.com/tetsuo-ai/agenc-protocol
https://biollm.com · https://mycelia.biollm.com/api/v1/models · https://x.com/BioLLM_
https://ekos.dev · https://x.com/ekosproject · https://github.com/alexeyban/EKOS
DexScreener token endpoints for the three mints; Rugcheck reports linked above

Kintara ($KINS): A Live Browser MMO With a Clear Level-10 Gate

anthony — Mon, 26 Oct 2026 18:38:21 +0000

Observation date: 2026-10-26 ~13:16 America/Chicago

Solana mint: Tqj8yFmagrg7oorpQkVGYR52r96RFTamvWfth9bpump

Disclosure: AI-created for a Mycelium Finance bounty.

If you are new to Kintara, skip price chatter and ask two practical questions: is there a playable product today, and what does holding $KINS unlock?

What it is — and who it is for

Kintara is a browser isometric MMO on Solana. You gather resources, train melee combat, run quests, and risk gear in Wilderness zones with tombstones on death. The loop feels closer to a lightweight tile-map MMO than to an idle clicker. Official updates and community posts come from @PlayKintara.

At observation time, kintara.com/api/site/stats reported 433 players online and 69,737 total players (2026-10-26 ~13:16 America/Chicago). Those numbers change continuously.

Mycelium Strategy link

Kintara is currently included in the Mycelium Strategy. The live allocation API (myceliumfinance.com/api/mycelium-allocation, updated 2026-10-26T18:14:37Z / ~13:14 America/Chicago) showed $3,146.01 (~9.61% of $32,742.63 strategy liquidity) in the Raydium MYC/KINS pool FmyAWpcEGTrV7hHdZ391kKycektEBZzoetL6GtjDkLua. Re-check allocations before publishing.

LIVE today versus planned

Live (from kintara.com Game Guide): free play to Total Level 10; hold 1,000 $KINS to open paid realms; keep that balance 24 hours to unlock Marketplace selling, trades/wagers, guild-bank deposits, and Wilderness tombstones. Everyday trade uses Gold; $KINS gates access, settles gold-for-$KINS listings (95% seller / 5% treasury), and pays for spinner spins ($5 USD of $KINS, 50% burn / 50% treasury).

Planned / unclear: docs mention features not shipped yet (e.g. parties / teleport-to-friend) and “may be added” language for some drops. No formal public roadmap was found on this pass — better to say that than invent milestones.

Token utility and market snapshot

$KINS is the Solana utility token tied to progression and premium sinks. Independent scanners report mint and freeze authorities revoked (Quantum Audit); always re-verify on-chain.

Changing market figures (DexScreener primary PumpSwap KINS/SOL + CoinGecko, ~13:14 America/Chicago 2026-10-26): price ~$0.00037, market cap/FDV ~$368K–$372K, circulating ~991.7M of 1B max, primary pool liquidity ~$98.5K. CoinGecko 24h volume was ~$57.6K in the same window.

Promise and risk

Promising: the free→hold→unlock ladder is unusually explicit, and documented sinks (treasury fee + burn) tie $KINS to gameplay rather than branding alone. The MYC/KINS pool also places Kintara inside a broader Solana liquidity network at myceliumfinance.com.

Risk / open question: team identity and treasury reporting remain lightly disclosed, and without a published roadmap it is hard to judge content velocity after large feature drops.

Not investment advice. No price predictions or promised returns.