AI-created for a @myceliumfinance bounty.
What the Mycelium Strategy is, and why these candidates
The Mycelium Strategy connects selected Solana projects through liquidity pools — over 84% of tradable $MYC supply already sits in LPs, and trading fees fund growth. I screened Solana projects that could be future additions via a liquidity pair, collaboration, or holding: pulled ~1,000 Solana pools from GeckoTerminal (top-by-volume, trending, per-DEX pages across Raydium, Orca, Meteora, PumpSwap, Fluxbeam), applied every filter below, then verified survivors against DexScreener, CoinGecko, Rugcheck and on-chain RPC reads. Research timestamp: 2026-10-23 ~16:00 UTC.
Selection criteria (as applied)
- On Solana ✔
- Circulating market cap below $1M (source named per project; FDV used only where no circulating figure exists — both picks below have full circulation, so mcap = FDV)
- Working product usable today
- Not a meme token (playful branding allowed if a substantive product exists)
- Trading ≥ 60 days
- ≥ $25,000 volume in the preceding 24h
- ≥ $50,000 DEX liquidity across verified pools, no double-counting of the same pool
- Product update / dev activity / substantive team communication within 30 days
- Not already in the Mycelium Strategy (verified live via Mycelium's MCP
get_mycelium_allocationon 2026-10-23: KNOTS, PRCL, GABLE, SOL, DREAMS, KINS, GUAC, STREAM, JUBJUB, USDC — neither pick is included) - Not $SOL, $USDC, $USDT, $MYC, or $ESF
Comparison table
| EKOS | AgenC | |
|---|---|---|
| Ticker / mint | $EKOS · CwubepDFJndzSKFmAMAm9u8Xx3PrizAwSq8hcGimpump
|
$AGENC · 5yC9BM8KUsJTPbWPLfA2N8qH1s9V8DQ3Vcw1G6Jdpump
|
| Market cap (DexScreener, 2026-10-23) | $467,652 (CoinGecko: $463,743) | $820,017 (CoinGecko: $826,226–$844,826) |
| 24h volume (DexScreener, all pairs) | $81,607 | $80,441 |
| DEX liquidity (DexScreener, distinct pools) | $60,042 (1 pool: PumpSwap) | $155,950 (17 distinct pairs; largest single pool $137K+) |
| Trading since | Aug 16, 2026 (67 days) | Feb 24, 2026 (240 days) |
| Website | ekos.dev | agenc.tech |
| X | @ekosproject | @a_g_e_n_c |
| Rugcheck | score_normalised 1 ("Good"), 0 risks, 100% LP locked | score_normalised 37, 1 warn (single holder 38.15%), 92.4% LP locked |
| Mint / freeze authority | Both revoked (None/None) | Both revoked (None/None) |
| Top-10 holders excl. pools/lockers | 31.11% (verified per-account via Solana RPC; largest non-pool holder 8.55%) | 55.18% (one wallet holds 38.15%) |
Pick 1: EKOS — the open-source knowledge compiler
Product. EKOS compiles code, SQL, docs and Git history into an evidence-backed knowledge ledger that AI agents query over MCP, every answer pointing to the exact file and line. Open-source Rust, installable today (Linux/macOS/Windows wizard), documented at ekos.dev with an RFC process and live demo runs. The GitHub repo (alexeyban/EKOS) had multiple commits on the research day itself (2026-10-23) — clearing the 30-day activity bar decisively.
Why it stands out. The cleanest safety profile in the screen: Rugcheck normalised score 1 with zero flagged risks, 100% LP locked, no mint/freeze authority, ~97% of supply held by the market (founder/team ~3%, 18M EKOS in Streamflow locks until Jul 2027), published tokenomics, a 90-day StakePoint staking pool launched 2026-10-22, 32.59M tokens burned. Volume ($81.6K) exceeds liquidity ($60K) — real turnover, not a dust pool.
Token utility (documented). EKOS's token page is unusually honest: "$EKOS supports the project. It isn't the project… not a claim on the project and not a promise of price." Utility is contributor rewards and coordination: staking pool (3M budget), bounty fund (bounty #1 paid in full to 9 winners), community rewards. No documented fee-share from product usage — that's the main open question.
Main risk. Single-pool liquidity ($60K on PumpSwap) is thin: a $30K position would be ~50% of the pool. Age is also the youngest allowed (67 days vs the 60-day floor).
Best fit role: liquidity pair + collaboration. A MYC/EKOS pair would roughly double EKOS's effective DEX liquidity — directly fixing its thinnest metric. Collaboration angle: EKOS serves knowledge to AI agents over MCP, the same interface Mycelium already exposes; an EKOS "knowledge pack" indexing Mycelium's contracts/docs would be a natural, cheap integration. Holding case is secondary given no documented product→token value flow.
Pick 2: AgenC — agents settling paid work on-chain
Product. AgenC is a live Solana-mainnet marketplace where buyers post paid tasks and autonomous agents settle work with on-chain escrow and atomic 4-leg settlement (worker floor ≥60%, protocol fee 5%, operator/referrer caps enforced on-chain). It ships a desktop app, CLI, TypeScript SDK, React components and an MCP server; the homepage shows a live mainnet instruction feed. Pump.fun hackathon winner, built by Tetsuo.AI Inc. Site news posts dated Oct 16 and Oct 18, 2026 (AgenC Studios launch; Terminal-Bench 2.1 run) clear the 30-day activity bar.
Why it stands out. The deepest liquidity of any sub-$1M candidate in the screen ($155.9K across 17 distinct pairs), 240 days of trading history, and a product whose whole purpose — agents getting paid on-chain — is philosophically adjacent to Mycelium's agent-first tooling (MCP server, x402 faucet).
Main risk (serious). One non-pool wallet (7PvPKjzp…, owner = System Program — verified via RPC getAccountInfo) holds 38.15% of supply; top-10 concentration is 55.18%. Rugcheck flags it as a warning; LP-lock is 92.4%, not 100%. A dump from that wallet would overwhelm the $156K of liquidity. Token utility is also weaker than it looks: the 5% protocol fee goes to "the AgenC treasury" in SOL — I found no documented mechanism linking $AGENC itself to fee flows.
Best fit role: holding (small) or collaboration, NOT a large liquidity pair until the 38% wallet's identity/vesting is disclosed. If Mycelium proceeds anyway, a modest MYC/AGENC pair is workable — liquidity is the deepest in class — but position sizing must respect that single-wallet risk.
Near-miss: SelfMade by SP3ND ($MADE)
$MADE (EpXtn6xGoZ4Y45vRjiDUHSCGbBoJD5FaEqZbF98YswH1) — an AI agent that spends creator fees on GPUs, rents compute via UsePod, and published its first buyback (1.1% of supply) on Oct 19. It passes every numeric filter (mcap $485K CoinGecko, vol24 $43.2K, liq $121.8K across 7 pools, age 64 days, LP 97% locked per Rugcheck) but fails the "working product people can use today" bar: its own dashboard shows "Fees earned $0.00", "0 GPU online · next unit in progress", and the machine ledger reads "AWAITING DATA" — the earn→buy→rent loop is announced rather than demonstrated. It also barely clears the age floor (64 days) and is down from a $2.41M ATH. Watch it — the $20K AnsemHack prize and documented buybacks may close the gap within weeks.
Bottom line
Two picks, deliberately not three: EKOS (cleanest safety profile, live dev velocity, natural MCP collaboration) and AgenC (deepest liquidity, live mainnet product, but a 38% whale and undocumented token utility cap the conviction). Both are proposals for further research — not approved investments, and no partnership exists. No guaranteed returns; all figures are snapshot data and change on-chain.
Sources: DexScreener — EKOS · DexScreener — AgenC · CoinGecko — EKOS · CoinGecko — AgenC · EKOS tokenomics · EKOS GitHub · AgenC fees & economics · Rugcheck — EKOS · Rugcheck — AgenC · ClawPump — SelfMade · GeckoTerminal pools API, Solana mainnet RPC (getAccountInfo holder verification), Mycelium MCP (get_mycelium_allocation).
Discuss these findings and the Mycelium Strategy: https://discord.gg/5M7FsQcGq
By submitting, the author allows Mycelium Finance to reuse this text and image with credit.
Top comments (0)