Solana Meme Coin Scanner Telemetry Confirms Live Execution Slippage and Router Latency

The transition of the Solana meme coin scanner from a simulated paper environment to live-money execution reveals severe slippage and router failures that mandate a complete halt of on-chain trading while paper-only data ingestion continues.

Live Mirror Discrepancies and Execution Halts

Live execution was halted at 14:08 BST on June 26, 2026, putting all on-chain trading systems into a fail-closed status. A manual halt was triggered via /root/Pumphouse/EXECUTION_PAUSED, leaving pumphouse-exec-mirror.service and pumphouse-mayhem-exec-mirror.service inactive. The halt followed significant discrepancies between simulated paper performance and funded wallet outcomes. For instance, paper replays for the Spike strategy showed simulated gains of +GBP 1.0235 on AbS2BZHWzauc4f6JQLYt5gtpx2Lbc7e9ZthJn8Jnpump and +GBP 1.0285 on another candidate, but the corresponding real-money execution wallet (gDQWTJHoeR4Fbka2HN1KJovUERmqt6dSTtU3ATSXeVc) produced net deltas of only +0.005295885 SOL (+GBP 0.2710) and +0.005315055 SOL (+GBP 0.2721). This real-money return of 31% to 33% contrasted with simulated paper figures exceeding 100%, revealing that standard latency models missed live routing dynamics. Paper-only tracking continues via pumphouse-listener.service.

Simulated Lane Performance and the Memecoin Paper Demo Ledger

The paper demo ledger compiles telemetry from six simulated trading lanes. Since the last reset to GBP 50.00, the model has processed events using a GBP 0.50 stake with momentum lanes at GBP 1.00. The BUNDLE lane led with simulated balances of GBP 466.38 (+416.38) on an 11.0% win rate over 2088 trades. The VELO lane showed GBP 183.66 (+133.66) across 189 trades (47.6% win rate), while the VELOR lane recorded GBP 155.38 (+105.38) over 360 trades (42.2% win rate). MEDIA_VELO reached GBP 73.16 (+23.16) on a 58.1% win rate over 31 trades, whereas CONFLUENCE_V drew down to GBP 41.12 (-8.88) on 1115 trades. Positions are logged in copy_demo_positions and visible on the Pumphouse dashboard.

Solana Holder Scanner and Indexing Metrics

Indexer freshness monitored by fc-system-spine shows the latest event is 0.0 minutes old, confirming an active chain connection. The 6-hour fresh board contains 200 tokens: 3 green, 172 amber, and 25 red. Counts in /root/Pumphouse/pump_events_live.db show the Solana holder scanner has processed 2,936,172 pump events, 3,501,771 early wallet rows, and 603,697 unique tokens. Raw events are parsed using the Solana RPC WebSocket protocol (WebSocket Docs) to map cluster risks. For long-term tracking, active timers on PumpSwap decoder and PumpSwap phase0 monitor token graduation rates and subsequent DEX survival.

Funder Intelligence and Pump.fun Cabal Detection

Cabal detection relies on creator funder mapping. Counts in /root/Pumphouse/funder_intel.db show 106,957 unique wallet funders, 186,122 funder events, and 71,578 scores. The system tracks 22,485 cross-token reused funders and 498 same-token multi-wallet funders. A systemd drop-in configuration on June 26, 2026, added 15 paper-only trigger funders to pumphouse-listener.service, including 2dbMWV..., Biw4ee..., FpwQQh..., and 4tTFM8.... Funder logs highlight fast-scaling tokens: GZMzT7RKqyJE1b7WwyoJz1WgkrQcsrwTU84J7pc2pump (from $1.6k to $132.7k in 68 seconds) and CKb9LF6oMATssc6Kk3hD2V99Tm9Jiy3WzqT1QJt4pump (from $2.0k to $132.0k in 66 seconds). Canonical token BcHEaaTCvycPwwsJ9yQTXdHP9X2gCLkznDbZ8VySpump shares funding signature G2YxRa6wt1qePMwfJzdXZG62ej4qaTC7YURzuh2Lwd3t.

Slippage Telemetry and Stale Quote Retries

Live tests revealed transaction routing bottlenecks. Token FRZezkKcd13ociXNfGA88oCQVB7wVSGbxvPx9VSZpump (cashcow) failed due to stale quote router error 6063. The operator patched /root/Pumphouse/services/pumphouse_exec_router.py to add a single retry factor of 0.35 in the buy call. Telemetry shows successful fills were highly relaxed: FomoPacks triggered at an $8k market cap but filled at $23k, while Human triggered at $3k but filled at $10k. To counter this drift, the system tightened PUMPHOUSE_EXEC_MIN_TOKEN_FACTOR from 0.35 to 0.65 and lowered PUMPHOUSE_EXEC_DRIFT_MAX to 1.50, enforcing strict transaction limits.

Exit Rule Research and Curve Stall Analysis

Exit-rule research has established a two-phase model for the experimental curve-stall lane. Telemetry from the June 6 winner 83mRwycNAmJLrBVg6mKQtcCTbuBM8qYigud6yoHApump (Jailed) showed the bot trimming 78.9% of the position and keeping a 21.1% runner. Consequently, /root/Pumphouse/services/pumphouse_mayhem_hot_paper.py was configured to hold a 0.210938 fraction at a $100k market cap, leaving the runner open until a 70% off-peak trailing stop triggers. To prevent active runners from locking capacity, PUMPHOUSE_EXEC_RUNNER_FREE_HELD_MAX was set to 0.25, exempting runners below this limit. These rules are verified on the Pumphouse method board using Pump.fun docs historical ticks.

System Boundaries and Operational Caveats

The local pump_events_live.db database starts on June 13, 2026, at 07:30:56, meaning raw transaction tapes from the June 5 and June 6 peaks are unavailable for latency replays. Replay outcomes on the Spike optimizer configuration remain negative: 18 replays returned a net loss of -GBP 6.41 and a 16.7% win rate. At the infrastructure level, /root/FreedomCore_Tools/bin/fc-system-spine marks the spine health as degraded due to observability and maverick alerts. Expanded funder mappings and slip parameters apply only to simulated profiles; live trading is paused until a new gateway is authorized. Research details are logged in the Atlas notes and hosted on FreedomCore.