• Joined on 2026-02-17
grenade pushed to main at swym/scout 2026-03-10 11:13:05 +00:00
d76d3b9061 Use write_all for ledger entries to improve concurrent-write safety
grenade pushed to main at swym/scout 2026-03-10 11:12:14 +00:00
0945c94cc8 Add --ledger-file arg for explicit ledger path control
grenade pushed to main at swym/scout 2026-03-10 11:05:46 +00:00
a0316be798 Add cross-run learning via run ledger and compare endpoint
grenade pushed to main at swym/scout 2026-03-10 11:04:23 +00:00
609d64587b docs: cross-run learnings plan
grenade pushed to main at swym/swym 2026-03-10 10:52:12 +00:00
4248bb0d1f docs: plan implemented by e47c184
e47c18404a feat: enrich backtest result_summary + add compare endpoint
Compare 2 commits »
grenade pushed to main at swym/scout 2026-03-10 10:25:08 +00:00
6692bdb490 Prompt: fix method vs kind confusion causing 11/15 validation failures
grenade pushed to main at swym/scout 2026-03-10 10:14:15 +00:00
36689e3fbb Prompt: fix field+offset kind omission and add interval guidance
grenade pushed to main at swym/scout 2026-03-10 09:43:11 +00:00
87d31f8d7e Use flat result_summary fields from swym patch 8fb410311
3892ab37c1 fix: parse actual result_summary structure (backtest_metadata + assets)
85896752f2 fix: ValidationError.path optional, correct position_quantity usage in prompts
ee260ea4d5 fix: parse flat result_summary structure per updated API doc
3f8d4de7fb feat: add declarative SizingMethod types from upstream schema
Compare 11 commits »
grenade pushed to main at swym/swym 2026-03-10 09:36:59 +00:00
8fb410311e feat: flatten result_summary + document sizing methods
4d6f307e57 docs: implementation planning
Compare 2 commits »
grenade pushed to main at swym/swym 2026-03-10 08:10:32 +00:00
ecbabf0f74 fix(validate): catch unknown fields in sizing methods via raw-JSON walk
grenade pushed to main at swym/swym 2026-03-10 08:02:41 +00:00
e6bd74fe8a fix: sell orders skipped when SizingMethod returns None for missing balance
grenade pushed to main at swym/swym 2026-03-10 07:29:37 +00:00
9e6ee7b1ea feat: POST /api/v1/strategies/validate — structured DSL validation endpoint
84989e2d9c feat: declarative position sizing methods in strategy DSL
494ce68e92 fix: dashboard balance toFixed crash + backfill range/style tweaks
e6d464948f docs+script: document backfill strategy and chunk by quarter
216729eb25 feat: database backup and restore scripts
Compare 6 commits »
grenade pushed to main at swym/scout 2026-03-09 16:46:49 +00:00
51e452b607 feat: discover max_output_tokens from server at startup
89f7ba66e0 feat: model-family-aware token budgets and prompt style
Compare 2 commits »
grenade pushed to main at swym/scout 2026-03-09 16:35:25 +00:00
6f4f864d28 fix: increase max_tokens to 8192 for R1 reasoning overhead
185cb4586e fix: strip R1 think blocks before JSON extraction
Compare 2 commits »
grenade pushed to main at swym/scout 2026-03-09 15:59:37 +00:00
b947f48b01 feat: client-side validation, cycling detection, quantity prompt fix
grenade pushed to main at swym/scout 2026-03-09 15:45:02 +00:00
e27aabae34 feat(agent): improve LLM feedback loop and convergence detection
grenade pushed to main at swym/scout 2026-03-09 12:22:37 +00:00
fb1145acae fix(swym): parse result_summary from actual API response structure
c7a2d65539 fix(prompts): forbid dynamic quantity expressions, require plain decimal string
292c101859 docs(prompts): add DSL expression kind reference and three working examples
Compare 3 commits »
grenade pushed to main at swym/scout 2026-03-09 10:58:58 +00:00
fc9b7e094a feat(agent): add strategy quality introspection
grenade pushed to main at swym/scout 2026-03-09 10:24:36 +00:00
deb28f6714 chore: local defaults
grenade pushed to main at swym/scout 2026-03-09 08:29:00 +00:00
b7aa458e40 feat(claude): add configurable API base URL via --anthropic-url