Inject Kraiken.sol (outstandingSupply, mint/burn mechanics) and Stake.sol
(snatch, withdrawal, KRK exclusion from floor denominator) into the red-team
agent prompt so agents can reason from actual source rather than guesses.
- red-team.sh: read SOL_KRAIKEN and SOL_STAKE from onchain/src/ alongside
the other six contracts already injected
- red-team-program.md: add ### Kraiken.sol and ### Stake.sol sections in the
Source Code reference block (after PriceOracle.sol)
- AGENTS.md: document the full list of injected contracts in a new
"Red-team Agent Context" section; both files are now listed as in-scope
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- red-team-sweep.sh: reset CROSS_PATTERNS_FILE at sweep start to prevent
stale patterns from prior invocations contaminating a fresh run
- red-team-sweep.sh: wrap pattern-extraction Python in set +e/set -e and
capture output so log() prefix is applied; move memory truncation outside
the if-block so it runs unconditionally even if Python fails
- red-team.sh: filter entries where candidate == current_candidate before
grouping, removing self-referential cross-candidate evidence
- red-team.sh: skip entries with empty pattern key (both pattern and
strategy fields empty) to prevent spurious bucket merging
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- red-team-sweep.sh: after each candidate completes, extract all memory
entries into /tmp/red-team-cross-patterns.jsonl (append), then clear
the raw memory file so the next candidate starts with a fresh state
- red-team.sh: define CROSS_PATTERNS_FILE; before building the prompt,
read the cross-patterns file and generate a "Cross-Candidate
Intelligence" section grouped by abstract op pattern — universal
patterns (broke 2+ candidates), candidate-specific wins, and patterns
that held everywhere — each annotated with optimizer profiles
- The new section is injected into the Claude prompt above the existing
Previous Findings block, satisfying all acceptance criteria
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- make_pattern: replace text.find('stake')/find('unstake') with
re.search(r'\bstake\b')/re.search(r'\bunstake\b') so 'stake' is never
found as a substring of 'unstake' (bug #1)
- make_pattern: track first-occurrence position of each op and sort by
position before building the sequence string, preserving actual
execution order instead of a hardcoded canonical order (bug #2)
- insight capture: track insight_pri on the current dict; only overwrite
stored insight when new match has strictly higher priority (lower index),
preventing a late 'because...' clause from silently replacing an earlier
'Key Insight:' capture (warning #3)
- run_num: compute max(run)+1 from JSON entries instead of wc -l so run
numbers stay monotonically increasing after memory trim (info #4)
- red-team-sweep.sh: also set adaptive flag when any r37-r40 register has
a variable-form assignment (r40 = uint256(someVar)), catching candidates
where only one branch uses constants (warning #5)
- red-team-sweep.sh: remove unnecessary 'import sys as _sys' in except
block; sys is already in scope (nit #6)
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- Add CANDIDATE_NAME and OPTIMIZER_PROFILE env vars to red-team.sh
(defaults to "unknown" for standalone runs)
- Update extract_memory Python: new fields candidate, optimizer_profile,
pattern (abstract op sequence via make_pattern()), and improved insight
extraction that also captures WHY explanations (because/since/due to)
- Update MEMORY_SECTION Python: entries now grouped by candidate;
universal patterns (DECREASED across multiple candidates) surfaced first
- Update prompt: add "Current Attack Target" table with candidate/profile,
optimizer parameter explanations (CI/AW/AS/DD behavioral impact),
Rule 9 requiring pattern+insight per strategy, updated report format
with Pattern/Insight fields and universal-pattern conclusion field
- Update red-team-sweep.sh: after inject, parse OptimizerV3Push3.sol for
r40/r39/r38/r37 constants to build OPTIMIZER_PROFILE string; pass
CANDIDATE_NAME and OPTIMIZER_PROFILE as env vars to red-team.sh
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
## What
- `tools/push3-transpiler/inject.sh` — shared transpile+inject logic used by both batch-eval and red-team-sweep
- `batch-eval.sh` — replaced inline 60-line Python block with `inject.sh` call
- `scripts/harb-evaluator/red-team-sweep.sh` — red-teams each kindergarten seed using existing `red-team.sh`, with random smoke test gate
## Why
Sweep script kept breaking because I rewrote the injection logic instead of reusing batch-eval's proven Python. Now there's one copy.
## Testing
- inject.sh tested manually on DO box with optimizer_v3 seed
- Smoke test picks random seed, injects + compiles before starting sweep
Co-authored-by: openhands <openhands@all-hands.dev>
Reviewed-on: https://codeberg.org/johba/harb/pulls/806
Reviewed-by: review_bot <review_bot@noreply.codeberg.org>
- Fix class-level NatSpec: use accurate wording (width computed from
anchorWidth param provided by Optimizer) instead of imprecise
LiquidityManager attribution
- Fix inline comment in _setAnchorPosition (same stale 1-100% claim)
- Update PRODUCT-TRUTH.md and ARCHITECTURE.md which had the same
incorrect 1-100% range claim
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Remove the misleading "(1-100% width)" range claim from the ANCHOR NatSpec.
Anchor width enforcement lives in LiquidityManager, not this abstract, so
the comment is replaced with a note pointing to where enforcement actually occurs.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Corrected run 7 note in manifest.jsonl: CI and DD values were inverted
(CI=20%, DD=0 → CI=0%, DD=20%) to match stack-pop semantics of the
push sequence 200000000000000000 153 200000000000000000 0.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- Change WARNING to explicitly state "legacy CID format ... migration not supported, skipping"
- Expand comment near the startswith('candidate_') guard to document the CID format
contract and explain why re-admission is intentionally out of scope (no surviving
generation_N.jsonl files from runs 1-6 exist in the repo)
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- Pass seed basename into the admission Python block as argv[7]
- Add \`note\` field to every new evolved entry: "Evolved from <seed> (run<N> gen<G>)"
- Add migration comment noting entries admitted before this fix may have note: null
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- Replace unquoted heredoc (shell-injection path) with a temp file: the
shell loop now appends tab-separated filename/score lines to a temp
file, which is passed as a plain path argument to the Python manifest-
rewrite block. Python reads only file contents, never executes shell-
expanded strings.
- Add early abort on fitness.sh exit code 2 (infra error: Anvil down,
missing tool). Iterating past an infra failure produces no useful
results; aborting immediately surfaces the real problem.
- Remove unused `os` import from the manifest-rewrite Python block.
- Fix inaccurate comment in evolve.sh --diverse-seeds sampling: the pool
sampler does a flat random shuffle with no fitness weighting; null-
fitness seeds are not "treated as 0" — they are sampled with equal
probability to any other seed.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- Add evaluate-seeds.sh: standalone script that reads manifest.jsonl,
finds every entry with fitness: null, runs fitness.sh against each
seed file, and atomically writes results back to manifest.jsonl.
Supports --dry-run to preview without evaluating.
- Add comment to --diverse-seeds sampling in evolve.sh documenting that
null-fitness seeds are included with effective_fitness=0 and that
evaluate-seeds.sh should be run to score them.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
