Commit Graph

4 Commits

Author SHA1 Message Date
travisBREAKS 89206107fa fix(bench): remove hardcoded credential paths from benchmark runners (#177)
The `_load_api_key()` function in longmemeval_bench.py and locomo_bench.py
searched for API keys in a fixed path (`~/.config/lu/keys.json`) using
personal key names (`anthropic_milla`, `anthropic_claude_code_main`).

This leaks internal infrastructure details into the public codebase and
trains contributors to store credentials in a non-standard location
rather than using the standard ANTHROPIC_API_KEY env var.

Simplified to: CLI flag > env var > empty string. Updated help text
and HYBRID_MODE.md docs to match.

Co-authored-by: Tadao <tadao@travisfixes.com>
2026-04-11 23:14:36 -07:00
travisBREAKS d8b2db696f fix(bench): remove global SSL verification bypass in convomem_bench (#176)
The module-level `ssl._create_default_https_context = ssl._create_unverified_context`
disables certificate verification for ALL urllib requests in the process,
not just the benchmark's HuggingFace downloads. This silently exposes
the benchmark runner to MITM attacks.

If a specific environment needs to skip verification (e.g. corporate proxy),
users can set `PYTHONHTTPSVERIFY=0` or pass a custom ssl context per-request
rather than globally patching the ssl module.

Co-authored-by: Tadao <tadao@travisfixes.com>
2026-04-11 23:14:12 -07:00
bensig 6d8c462219 fix: resolve ruff lint and format errors across codebase
Fix E402 import ordering, F841 unused variable, F541 unnecessary
f-strings, F401 unused import, and auto-format 6 files.
2026-04-04 18:37:17 -07:00
bensig 0f8fa8c7d5 bench: add benchmark runners, results docs, and test suite
Benchmarks: LongMemEval, LoCoMo, ConvoMem, MemBench runners with
methodology docs and hybrid retrieval analysis.

Tests: config, miner, convo_miner, normalize — 9 tests, all passing.
2026-04-04 18:33:42 -07:00