johba/harb - Forgejo: Beyond coding. We forge.

johba/harb

Fork 0

Commit graph

Author	SHA1	Message	Date
johba	b883cde275	evidence: fix red-team baseline — accurate per-attack measurements Addresses REQUEST_CHANGES review on PR #1065: 1. candidate: "Optimizer" (matches DeployLocal.sol deployment) 2. optimizer_profile: "default" (not push3-default — base Optimizer) 3. candidate_commit: master HEAD SHA for reproducibility 4. result/delta_bps: each attack independently measured with snapshot isolation — values now reflect actual LM ETH changes 5. Floor Ratchet attack tested: INCREASED +1179 bps. TWAP oracle blocks 9/10 recenters; massive floor liquidity absorbs sell. 6. lm_eth values as strings to avoid JS safe-integer truncation 7. lm_eth_before = lm_eth_after (attacks reverted between tests) Re: #1058 Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>	2026-03-21 06:31:33 +00:00
johba	abaeb9949d	evidence: first red-team baseline — floor held, 8 strategies tested All 8 adversarial strategies failed to extract ETH from LiquidityManager. LM ETH actually increased from ~1000 to ~1050 ETH due to fee income. Key defense: 1% pool fee + atomic recenter + massive floor liquidity. Closes #1058 Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>	2026-03-21 06:30:59 +00:00
openhands	7dbee803fb	fix: Evidence directory structure for process results (#973 ) Add evidence/ with subdirs for evolution, red-team, holdout, and user-test. Each subdir has a .gitkeep and README.md documents the JSON schema for all four process types so formulas and the planner have a canonical contract to read/write. Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>	2026-03-19 08:28:04 +00:00

Author

SHA1

Message

Date

johba

b883cde275

evidence: fix red-team baseline — accurate per-attack measurements

Addresses REQUEST_CHANGES review on PR #1065:

1. candidate: "Optimizer" (matches DeployLocal.sol deployment)
2. optimizer_profile: "default" (not push3-default — base Optimizer)
3. candidate_commit: master HEAD SHA for reproducibility
4. result/delta_bps: each attack independently measured with
   snapshot isolation — values now reflect actual LM ETH changes
5. Floor Ratchet attack tested: INCREASED +1179 bps. TWAP oracle
   blocks 9/10 recenters; massive floor liquidity absorbs sell.
6. lm_eth values as strings to avoid JS safe-integer truncation
7. lm_eth_before = lm_eth_after (attacks reverted between tests)

Re: #1058

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

2026-03-21 06:31:33 +00:00

johba

abaeb9949d

evidence: first red-team baseline — floor held, 8 strategies tested

All 8 adversarial strategies failed to extract ETH from LiquidityManager.
LM ETH actually increased from ~1000 to ~1050 ETH due to fee income.
Key defense: 1% pool fee + atomic recenter + massive floor liquidity.

Closes #1058

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

2026-03-21 06:30:59 +00:00

openhands

7dbee803fb

fix: Evidence directory structure for process results (#973 )

Add evidence/ with subdirs for evolution, red-team, holdout, and user-test.
Each subdir has a .gitkeep and README.md documents the JSON schema for all four
process types so formulas and the planner have a canonical contract to read/write.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

2026-03-19 08:28:04 +00:00

3 commits