See which declared accounts connect transactions.

REAL NETWORK DATA — CAPTURED DEVNET FIXTURE

📤Real devnet: loader buffer writes

Explore a captured program upload and the declared locks shared by its decoded buffer writes. The report also compares hypothetical payer and byte-range changes.

Captured slot: 488638500 · Provenance: real-network / devnet · Retrieved 2026-08-27T17:58:43.489Z

Transactions

45

Declared conflict pairs

116

Cross-payer shared-account pairs

0

Estimated execution waves

15

Full-block wave bounds: 15–15. Known Compute Units coverage: 45/45 transactions. Compute Units measure work, not elapsed time.

What the analyzer found

• program-buffer writes (high confidence)

  This finding contains 15 transactions and 105 unique conflict pairs (90.5% of block pairs). It has one declared fee payer. The clique lower bound and generated-wave upper bound both equal 15, giving a proven minimum wave count of 15.

  Actor pattern: same payer. Actionability: likely protocol-required or intentional.

  Evidence behind this finding

  • 1 decoded BPF Upgradeable Loader write sequence targets shared conflict accounts in this component.
  • The RPC-decoded write fields identify both buffer and authority.
  • Contiguous writes to one decoded upgradeable-loader buffer form an ordered upload sequence.

📤Open interactive reportView analysis JSON

Shipped input: fixtures/real-devnet-block.json

REAL NETWORK DATA — CAPTURED DEVNET FIXTURE

🤝Real devnet: cross-payer funding

Find transactions with different fee payers that draw from a shared funding source. Different payers alone do not establish independent application work or a safe redesign.

Captured slot: 488638508 · Provenance: real-network / devnet · Retrieved 2026-08-27T17:58:27.792Z

Transactions

51

Declared conflict pairs

146

Cross-payer shared-account pairs

1

Estimated execution waves

17

Full-block wave bounds: 17–17. Known Compute Units coverage: 51/51 transactions. Compute Units measure work, not elapsed time.

What the analyzer found

• transfer/funding source (high confidence)

  This finding contains 2 transactions and 1 unique conflict pair (0.7% of block pairs). It spans 2 declared fee payers. The clique lower bound and generated-wave upper bound both equal 2, giving a proven minimum wave count of 2.

  Actor pattern: cross payer. Actionability: insufficient evidence.

  Evidence behind this finding

  • 2 RPC-decoded System Program transfers use the same writable source account 6yvwhesLJeE8fNWviosRoUtBP3VFUXE7SEhSP9fFRJ3Z.
  • Each decoded transfer specifies 20,000,000 lamports.
  • The decoded destinations differ.
  • Participating transaction outcomes: 2 successful, 0 failed, 0 unknown. Decoded amounts describe instructions, not verified balance changes.
  • Both participating transactions contain an Airdrop log.
  • Decoded evidence does not prove a safe alternative account layout or operation design.

🤝Open interactive reportView analysis JSON

Shipped input: fixtures/cross-payer-devnet-block.json

SYNTHETIC TEACHING FIXTURE — NOT OBSERVED NETWORK ACTIVITY

🧪Synthetic teaching: shared application state

A deliberately constructed example illustrates application-state contention and supplied evidence for an alternative partitioning. It is not evidence of an observed network opportunity.

Teaching slot identifier: 1500000 · Provenance: curated-teaching / synthetic

Transactions

4

Declared conflict pairs

3

Cross-payer shared-account pairs

3

Estimated execution waves

3

Full-block wave bounds: 3–3. Known Compute Units coverage: 4/4 transactions. Compute Units measure work, not elapsed time.

What the analyzer found

• application state (high confidence)

  This finding contains 3 transactions and 3 unique conflict pairs (100.0% of block pairs). It spans 3 declared fee payers. The clique lower bound and generated-wave upper bound both equal 3, giving a proven minimum wave count of 3.

  Actor pattern: cross payer. Actionability: insufficient evidence.

  Evidence behind this finding

  • RPC-decoded instruction fields identify a shared application-state account.
  • Decoding application state does not prove logical independence or provide a concrete alternative partitioning.

🧪Open interactive reportView analysis JSON

Shipped input: fixtures/cross-payer-application-state-teaching.json

REAL NETWORK DATA — CAPTURED DEVNET FIXTURES

🧵Real devnet: upload across nine blocks

Follow decoded payload chunks across slots. Inspect byte coverage, gaps, overlaps, transaction outcomes, and exact instruction evidence. Byte coverage does not establish final buffer contents or a complete upload lifecycle.

Captured slots: 488638500–488638508

Observed blocks

9

Decoded chunks

148

Upload sequences

1

🧵Open multi-block reportView range analysis JSON

Shipped inputs: fixtures/loader-range/

⚖️Compare two captured blocks

Compare the upload and funding examples side by side: transaction outcomes, declared conflict pairs, connected components, wave bounds, and recorded compute. These are different workloads, so their differences do not establish a performance improvement.

⚖️Open block comparisonView comparison JSON⚖️Compare your own files